diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,9597 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 3.0,
+  "eval_steps": 500,
+  "global_step": 1365,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.002197802197802198,
+      "grad_norm": 0.0018604777287691832,
+      "learning_rate": 0.0,
+      "loss": 0.904,
+      "step": 1
+    },
+    {
+      "epoch": 0.004395604395604396,
+      "grad_norm": 0.0022015220019966364,
+      "learning_rate": 2.18978102189781e-05,
+      "loss": 0.9679,
+      "step": 2
+    },
+    {
+      "epoch": 0.006593406593406593,
+      "grad_norm": 0.0038708504289388657,
+      "learning_rate": 4.37956204379562e-05,
+      "loss": 1.2015,
+      "step": 3
+    },
+    {
+      "epoch": 0.008791208791208791,
+      "grad_norm": 0.002634151605889201,
+      "learning_rate": 6.569343065693432e-05,
+      "loss": 1.1197,
+      "step": 4
+    },
+    {
+      "epoch": 0.01098901098901099,
+      "grad_norm": 0.002863324712961912,
+      "learning_rate": 8.75912408759124e-05,
+      "loss": 1.0253,
+      "step": 5
+    },
+    {
+      "epoch": 0.013186813186813187,
+      "grad_norm": 0.002474637469276786,
+      "learning_rate": 0.00010948905109489051,
+      "loss": 0.7961,
+      "step": 6
+    },
+    {
+      "epoch": 0.015384615384615385,
+      "grad_norm": 0.0018731632735580206,
+      "learning_rate": 0.00013138686131386864,
+      "loss": 0.8539,
+      "step": 7
+    },
+    {
+      "epoch": 0.017582417582417582,
+      "grad_norm": 0.0028935405425727367,
+      "learning_rate": 0.00015328467153284672,
+      "loss": 1.0031,
+      "step": 8
+    },
+    {
+      "epoch": 0.01978021978021978,
+      "grad_norm": 0.004036034923046827,
+      "learning_rate": 0.0001751824817518248,
+      "loss": 0.9504,
+      "step": 9
+    },
+    {
+      "epoch": 0.02197802197802198,
+      "grad_norm": 0.003619411261752248,
+      "learning_rate": 0.00019708029197080293,
+      "loss": 1.0529,
+      "step": 10
+    },
+    {
+      "epoch": 0.024175824175824177,
+      "grad_norm": 0.0018631212878972292,
+      "learning_rate": 0.00021897810218978101,
+      "loss": 0.9699,
+      "step": 11
+    },
+    {
+      "epoch": 0.026373626373626374,
+      "grad_norm": 0.0030637774616479874,
+      "learning_rate": 0.00024087591240875913,
+      "loss": 1.0136,
+      "step": 12
+    },
+    {
+      "epoch": 0.02857142857142857,
+      "grad_norm": 0.0018094092374667525,
+      "learning_rate": 0.00026277372262773727,
+      "loss": 0.8956,
+      "step": 13
+    },
+    {
+      "epoch": 0.03076923076923077,
+      "grad_norm": 0.0021992847323417664,
+      "learning_rate": 0.00028467153284671533,
+      "loss": 0.9303,
+      "step": 14
+    },
+    {
+      "epoch": 0.03296703296703297,
+      "grad_norm": 0.0045089018531143665,
+      "learning_rate": 0.00030656934306569345,
+      "loss": 1.3117,
+      "step": 15
+    },
+    {
+      "epoch": 0.035164835164835165,
+      "grad_norm": 0.0020882843527942896,
+      "learning_rate": 0.00032846715328467156,
+      "loss": 0.8858,
+      "step": 16
+    },
+    {
+      "epoch": 0.03736263736263736,
+      "grad_norm": 0.0020202661398798227,
+      "learning_rate": 0.0003503649635036496,
+      "loss": 1.0535,
+      "step": 17
+    },
+    {
+      "epoch": 0.03956043956043956,
+      "grad_norm": 0.0022548639681190252,
+      "learning_rate": 0.00037226277372262774,
+      "loss": 1.0383,
+      "step": 18
+    },
+    {
+      "epoch": 0.041758241758241756,
+      "grad_norm": 0.002697764430195093,
+      "learning_rate": 0.00039416058394160585,
+      "loss": 1.061,
+      "step": 19
+    },
+    {
+      "epoch": 0.04395604395604396,
+      "grad_norm": 0.004127498250454664,
+      "learning_rate": 0.00041605839416058397,
+      "loss": 1.1928,
+      "step": 20
+    },
+    {
+      "epoch": 0.046153846153846156,
+      "grad_norm": 0.003866023849695921,
+      "learning_rate": 0.00043795620437956203,
+      "loss": 0.8339,
+      "step": 21
+    },
+    {
+      "epoch": 0.04835164835164835,
+      "grad_norm": 0.005134144797921181,
+      "learning_rate": 0.0004598540145985402,
+      "loss": 0.9898,
+      "step": 22
+    },
+    {
+      "epoch": 0.05054945054945055,
+      "grad_norm": 0.003004892263561487,
+      "learning_rate": 0.00048175182481751826,
+      "loss": 1.0775,
+      "step": 23
+    },
+    {
+      "epoch": 0.05274725274725275,
+      "grad_norm": 0.002847501542419195,
+      "learning_rate": 0.0005036496350364964,
+      "loss": 1.026,
+      "step": 24
+    },
+    {
+      "epoch": 0.054945054945054944,
+      "grad_norm": 0.0018494039541110396,
+      "learning_rate": 0.0005255474452554745,
+      "loss": 0.9018,
+      "step": 25
+    },
+    {
+      "epoch": 0.05714285714285714,
+      "grad_norm": 0.002604322973638773,
+      "learning_rate": 0.0005474452554744526,
+      "loss": 0.8395,
+      "step": 26
+    },
+    {
+      "epoch": 0.05934065934065934,
+      "grad_norm": 0.0025429942179471254,
+      "learning_rate": 0.0005693430656934307,
+      "loss": 1.0386,
+      "step": 27
+    },
+    {
+      "epoch": 0.06153846153846154,
+      "grad_norm": 0.001954495208337903,
+      "learning_rate": 0.0005912408759124088,
+      "loss": 0.922,
+      "step": 28
+    },
+    {
+      "epoch": 0.06373626373626373,
+      "grad_norm": 0.0031225685961544514,
+      "learning_rate": 0.0006131386861313869,
+      "loss": 1.1752,
+      "step": 29
+    },
+    {
+      "epoch": 0.06593406593406594,
+      "grad_norm": 0.0019849054515361786,
+      "learning_rate": 0.000635036496350365,
+      "loss": 1.0344,
+      "step": 30
+    },
+    {
+      "epoch": 0.06813186813186813,
+      "grad_norm": 0.0025169739965349436,
+      "learning_rate": 0.0006569343065693431,
+      "loss": 0.955,
+      "step": 31
+    },
+    {
+      "epoch": 0.07032967032967033,
+      "grad_norm": 0.0024823860730975866,
+      "learning_rate": 0.0006788321167883212,
+      "loss": 1.1081,
+      "step": 32
+    },
+    {
+      "epoch": 0.07252747252747253,
+      "grad_norm": 0.003076680935919285,
+      "learning_rate": 0.0007007299270072992,
+      "loss": 1.2123,
+      "step": 33
+    },
+    {
+      "epoch": 0.07472527472527472,
+      "grad_norm": 0.001619118032976985,
+      "learning_rate": 0.0007226277372262774,
+      "loss": 0.9891,
+      "step": 34
+    },
+    {
+      "epoch": 0.07692307692307693,
+      "grad_norm": 0.0020884708501398563,
+      "learning_rate": 0.0007445255474452555,
+      "loss": 0.934,
+      "step": 35
+    },
+    {
+      "epoch": 0.07912087912087912,
+      "grad_norm": 0.002584906993433833,
+      "learning_rate": 0.0007664233576642336,
+      "loss": 1.0778,
+      "step": 36
+    },
+    {
+      "epoch": 0.08131868131868132,
+      "grad_norm": 0.005661378148943186,
+      "learning_rate": 0.0007883211678832117,
+      "loss": 0.9046,
+      "step": 37
+    },
+    {
+      "epoch": 0.08351648351648351,
+      "grad_norm": 0.002825263887643814,
+      "learning_rate": 0.0008102189781021898,
+      "loss": 0.9346,
+      "step": 38
+    },
+    {
+      "epoch": 0.08571428571428572,
+      "grad_norm": 0.0022831654641777277,
+      "learning_rate": 0.0008321167883211679,
+      "loss": 0.9237,
+      "step": 39
+    },
+    {
+      "epoch": 0.08791208791208792,
+      "grad_norm": 0.0044238674454391,
+      "learning_rate": 0.000854014598540146,
+      "loss": 1.1424,
+      "step": 40
+    },
+    {
+      "epoch": 0.09010989010989011,
+      "grad_norm": 0.0022852844558656216,
+      "learning_rate": 0.0008759124087591241,
+      "loss": 0.8855,
+      "step": 41
+    },
+    {
+      "epoch": 0.09230769230769231,
+      "grad_norm": 0.004308356903493404,
+      "learning_rate": 0.0008978102189781023,
+      "loss": 0.9858,
+      "step": 42
+    },
+    {
+      "epoch": 0.0945054945054945,
+      "grad_norm": 0.00387572031468153,
+      "learning_rate": 0.0009197080291970804,
+      "loss": 1.1803,
+      "step": 43
+    },
+    {
+      "epoch": 0.0967032967032967,
+      "grad_norm": 0.0018792622722685337,
+      "learning_rate": 0.0009416058394160585,
+      "loss": 0.8865,
+      "step": 44
+    },
+    {
+      "epoch": 0.0989010989010989,
+      "grad_norm": 0.0020905013661831617,
+      "learning_rate": 0.0009635036496350365,
+      "loss": 0.9194,
+      "step": 45
+    },
+    {
+      "epoch": 0.1010989010989011,
+      "grad_norm": 0.003089776262640953,
+      "learning_rate": 0.0009854014598540146,
+      "loss": 0.8324,
+      "step": 46
+    },
+    {
+      "epoch": 0.10329670329670329,
+      "grad_norm": 0.002898598089814186,
+      "learning_rate": 0.0010072992700729927,
+      "loss": 0.8711,
+      "step": 47
+    },
+    {
+      "epoch": 0.1054945054945055,
+      "grad_norm": 0.002743329619988799,
+      "learning_rate": 0.001029197080291971,
+      "loss": 1.2497,
+      "step": 48
+    },
+    {
+      "epoch": 0.1076923076923077,
+      "grad_norm": 0.002174532040953636,
+      "learning_rate": 0.001051094890510949,
+      "loss": 0.9713,
+      "step": 49
+    },
+    {
+      "epoch": 0.10989010989010989,
+      "grad_norm": 0.002874185098335147,
+      "learning_rate": 0.001072992700729927,
+      "loss": 0.9073,
+      "step": 50
+    },
+    {
+      "epoch": 0.11208791208791209,
+      "grad_norm": 0.0037335872184485197,
+      "learning_rate": 0.0010948905109489052,
+      "loss": 1.1499,
+      "step": 51
+    },
+    {
+      "epoch": 0.11428571428571428,
+      "grad_norm": 0.007152085192501545,
+      "learning_rate": 0.0011167883211678832,
+      "loss": 1.1822,
+      "step": 52
+    },
+    {
+      "epoch": 0.11648351648351649,
+      "grad_norm": 0.009064623154699802,
+      "learning_rate": 0.0011386861313868613,
+      "loss": 1.3917,
+      "step": 53
+    },
+    {
+      "epoch": 0.11868131868131868,
+      "grad_norm": 0.0029882113449275494,
+      "learning_rate": 0.0011605839416058395,
+      "loss": 1.1625,
+      "step": 54
+    },
+    {
+      "epoch": 0.12087912087912088,
+      "grad_norm": 0.0033028488978743553,
+      "learning_rate": 0.0011824817518248177,
+      "loss": 0.9634,
+      "step": 55
+    },
+    {
+      "epoch": 0.12307692307692308,
+      "grad_norm": 0.004085292108356953,
+      "learning_rate": 0.0012043795620437956,
+      "loss": 0.8678,
+      "step": 56
+    },
+    {
+      "epoch": 0.12527472527472527,
+      "grad_norm": 0.003470406401902437,
+      "learning_rate": 0.0012262773722627738,
+      "loss": 0.9782,
+      "step": 57
+    },
+    {
+      "epoch": 0.12747252747252746,
+      "grad_norm": 0.002610692521557212,
+      "learning_rate": 0.0012481751824817517,
+      "loss": 1.0321,
+      "step": 58
+    },
+    {
+      "epoch": 0.12967032967032968,
+      "grad_norm": 0.003164509776979685,
+      "learning_rate": 0.00127007299270073,
+      "loss": 1.1286,
+      "step": 59
+    },
+    {
+      "epoch": 0.13186813186813187,
+      "grad_norm": 0.0019173154141753912,
+      "learning_rate": 0.001291970802919708,
+      "loss": 1.0541,
+      "step": 60
+    },
+    {
+      "epoch": 0.13406593406593406,
+      "grad_norm": 0.0027315299957990646,
+      "learning_rate": 0.0013138686131386862,
+      "loss": 1.0978,
+      "step": 61
+    },
+    {
+      "epoch": 0.13626373626373625,
+      "grad_norm": 0.013589787296950817,
+      "learning_rate": 0.0013357664233576642,
+      "loss": 1.1554,
+      "step": 62
+    },
+    {
+      "epoch": 0.13846153846153847,
+      "grad_norm": 0.005177911836653948,
+      "learning_rate": 0.0013576642335766424,
+      "loss": 0.7935,
+      "step": 63
+    },
+    {
+      "epoch": 0.14065934065934066,
+      "grad_norm": 0.0027772269677370787,
+      "learning_rate": 0.0013795620437956205,
+      "loss": 1.0377,
+      "step": 64
+    },
+    {
+      "epoch": 0.14285714285714285,
+      "grad_norm": 0.0031254368368536234,
+      "learning_rate": 0.0014014598540145985,
+      "loss": 1.0697,
+      "step": 65
+    },
+    {
+      "epoch": 0.14505494505494507,
+      "grad_norm": 0.002790456172078848,
+      "learning_rate": 0.0014233576642335767,
+      "loss": 0.9814,
+      "step": 66
+    },
+    {
+      "epoch": 0.14725274725274726,
+      "grad_norm": 0.0023073949851095676,
+      "learning_rate": 0.0014452554744525548,
+      "loss": 1.0949,
+      "step": 67
+    },
+    {
+      "epoch": 0.14945054945054945,
+      "grad_norm": 0.002217709319666028,
+      "learning_rate": 0.001467153284671533,
+      "loss": 0.9828,
+      "step": 68
+    },
+    {
+      "epoch": 0.15164835164835164,
+      "grad_norm": 0.004850276745855808,
+      "learning_rate": 0.001489051094890511,
+      "loss": 0.9461,
+      "step": 69
+    },
+    {
+      "epoch": 0.15384615384615385,
+      "grad_norm": 0.003828238695859909,
+      "learning_rate": 0.0015109489051094893,
+      "loss": 1.1597,
+      "step": 70
+    },
+    {
+      "epoch": 0.15604395604395604,
+      "grad_norm": 0.0025202587712556124,
+      "learning_rate": 0.0015328467153284673,
+      "loss": 1.1554,
+      "step": 71
+    },
+    {
+      "epoch": 0.15824175824175823,
+      "grad_norm": 0.006070408038794994,
+      "learning_rate": 0.0015547445255474455,
+      "loss": 0.9586,
+      "step": 72
+    },
+    {
+      "epoch": 0.16043956043956045,
+      "grad_norm": 0.0038373274728655815,
+      "learning_rate": 0.0015766423357664234,
+      "loss": 1.2333,
+      "step": 73
+    },
+    {
+      "epoch": 0.16263736263736264,
+      "grad_norm": 0.0039412518963217735,
+      "learning_rate": 0.0015985401459854016,
+      "loss": 1.1171,
+      "step": 74
+    },
+    {
+      "epoch": 0.16483516483516483,
+      "grad_norm": 0.002648682100698352,
+      "learning_rate": 0.0016204379562043795,
+      "loss": 1.0811,
+      "step": 75
+    },
+    {
+      "epoch": 0.16703296703296702,
+      "grad_norm": 0.004417882300913334,
+      "learning_rate": 0.0016423357664233577,
+      "loss": 0.8881,
+      "step": 76
+    },
+    {
+      "epoch": 0.16923076923076924,
+      "grad_norm": 0.0036130344960838556,
+      "learning_rate": 0.0016642335766423359,
+      "loss": 0.9779,
+      "step": 77
+    },
+    {
+      "epoch": 0.17142857142857143,
+      "grad_norm": 0.003095177235081792,
+      "learning_rate": 0.0016861313868613138,
+      "loss": 0.9355,
+      "step": 78
+    },
+    {
+      "epoch": 0.17362637362637362,
+      "grad_norm": 0.0036904679145663977,
+      "learning_rate": 0.001708029197080292,
+      "loss": 0.9673,
+      "step": 79
+    },
+    {
+      "epoch": 0.17582417582417584,
+      "grad_norm": 0.0035005153622478247,
+      "learning_rate": 0.00172992700729927,
+      "loss": 1.0764,
+      "step": 80
+    },
+    {
+      "epoch": 0.17802197802197803,
+      "grad_norm": 0.002090970752760768,
+      "learning_rate": 0.0017518248175182481,
+      "loss": 0.9784,
+      "step": 81
+    },
+    {
+      "epoch": 0.18021978021978022,
+      "grad_norm": 0.00546365138143301,
+      "learning_rate": 0.0017737226277372265,
+      "loss": 0.9245,
+      "step": 82
+    },
+    {
+      "epoch": 0.1824175824175824,
+      "grad_norm": 0.003848861902952194,
+      "learning_rate": 0.0017956204379562047,
+      "loss": 1.1459,
+      "step": 83
+    },
+    {
+      "epoch": 0.18461538461538463,
+      "grad_norm": 0.0033583431504666805,
+      "learning_rate": 0.0018175182481751826,
+      "loss": 1.1118,
+      "step": 84
+    },
+    {
+      "epoch": 0.18681318681318682,
+      "grad_norm": 0.003119210246950388,
+      "learning_rate": 0.0018394160583941608,
+      "loss": 1.0026,
+      "step": 85
+    },
+    {
+      "epoch": 0.189010989010989,
+      "grad_norm": 0.0034774355590343475,
+      "learning_rate": 0.0018613138686131387,
+      "loss": 1.169,
+      "step": 86
+    },
+    {
+      "epoch": 0.1912087912087912,
+      "grad_norm": 0.0031014415435492992,
+      "learning_rate": 0.001883211678832117,
+      "loss": 0.7906,
+      "step": 87
+    },
+    {
+      "epoch": 0.1934065934065934,
+      "grad_norm": 0.003485433990135789,
+      "learning_rate": 0.0019051094890510949,
+      "loss": 0.9413,
+      "step": 88
+    },
+    {
+      "epoch": 0.1956043956043956,
+      "grad_norm": 0.0023805852979421616,
+      "learning_rate": 0.001927007299270073,
+      "loss": 1.0094,
+      "step": 89
+    },
+    {
+      "epoch": 0.1978021978021978,
+      "grad_norm": 0.002914723241701722,
+      "learning_rate": 0.001948905109489051,
+      "loss": 1.0001,
+      "step": 90
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.0023641674779355526,
+      "learning_rate": 0.001970802919708029,
+      "loss": 0.9358,
+      "step": 91
+    },
+    {
+      "epoch": 0.2021978021978022,
+      "grad_norm": 0.0033865005243569613,
+      "learning_rate": 0.0019927007299270073,
+      "loss": 0.7391,
+      "step": 92
+    },
+    {
+      "epoch": 0.2043956043956044,
+      "grad_norm": 0.0037255098577588797,
+      "learning_rate": 0.0020145985401459855,
+      "loss": 1.0169,
+      "step": 93
+    },
+    {
+      "epoch": 0.20659340659340658,
+      "grad_norm": 0.003881957847625017,
+      "learning_rate": 0.0020364963503649632,
+      "loss": 1.0658,
+      "step": 94
+    },
+    {
+      "epoch": 0.2087912087912088,
+      "grad_norm": 0.004074831027537584,
+      "learning_rate": 0.002058394160583942,
+      "loss": 0.9837,
+      "step": 95
+    },
+    {
+      "epoch": 0.210989010989011,
+      "grad_norm": 0.005445981863886118,
+      "learning_rate": 0.00208029197080292,
+      "loss": 1.1292,
+      "step": 96
+    },
+    {
+      "epoch": 0.21318681318681318,
+      "grad_norm": 0.005073125474154949,
+      "learning_rate": 0.002102189781021898,
+      "loss": 0.9083,
+      "step": 97
+    },
+    {
+      "epoch": 0.2153846153846154,
+      "grad_norm": 0.002276445971801877,
+      "learning_rate": 0.002124087591240876,
+      "loss": 1.2016,
+      "step": 98
+    },
+    {
+      "epoch": 0.2175824175824176,
+      "grad_norm": 0.003270257031545043,
+      "learning_rate": 0.002145985401459854,
+      "loss": 1.2341,
+      "step": 99
+    },
+    {
+      "epoch": 0.21978021978021978,
+      "grad_norm": 0.0036891719792038202,
+      "learning_rate": 0.0021678832116788322,
+      "loss": 1.0599,
+      "step": 100
+    },
+    {
+      "epoch": 0.22197802197802197,
+      "grad_norm": 0.002439539646729827,
+      "learning_rate": 0.0021897810218978104,
+      "loss": 1.0029,
+      "step": 101
+    },
+    {
+      "epoch": 0.22417582417582418,
+      "grad_norm": 0.006534152664244175,
+      "learning_rate": 0.002211678832116788,
+      "loss": 0.9815,
+      "step": 102
+    },
+    {
+      "epoch": 0.22637362637362637,
+      "grad_norm": 0.003772962838411331,
+      "learning_rate": 0.0022335766423357663,
+      "loss": 0.8686,
+      "step": 103
+    },
+    {
+      "epoch": 0.22857142857142856,
+      "grad_norm": 0.0035889260470867157,
+      "learning_rate": 0.0022554744525547445,
+      "loss": 0.9956,
+      "step": 104
+    },
+    {
+      "epoch": 0.23076923076923078,
+      "grad_norm": 0.0036207009106874466,
+      "learning_rate": 0.0022773722627737226,
+      "loss": 1.1197,
+      "step": 105
+    },
+    {
+      "epoch": 0.23296703296703297,
+      "grad_norm": 0.005692319478839636,
+      "learning_rate": 0.002299270072992701,
+      "loss": 0.996,
+      "step": 106
+    },
+    {
+      "epoch": 0.23516483516483516,
+      "grad_norm": 0.002964536426588893,
+      "learning_rate": 0.002321167883211679,
+      "loss": 1.0277,
+      "step": 107
+    },
+    {
+      "epoch": 0.23736263736263735,
+      "grad_norm": 0.00476460624486208,
+      "learning_rate": 0.002343065693430657,
+      "loss": 1.0381,
+      "step": 108
+    },
+    {
+      "epoch": 0.23956043956043957,
+      "grad_norm": 0.0026263007894158363,
+      "learning_rate": 0.0023649635036496353,
+      "loss": 0.9696,
+      "step": 109
+    },
+    {
+      "epoch": 0.24175824175824176,
+      "grad_norm": 0.0052644032984972,
+      "learning_rate": 0.002386861313868613,
+      "loss": 1.0141,
+      "step": 110
+    },
+    {
+      "epoch": 0.24395604395604395,
+      "grad_norm": 0.004063093103468418,
+      "learning_rate": 0.0024087591240875912,
+      "loss": 1.0943,
+      "step": 111
+    },
+    {
+      "epoch": 0.24615384615384617,
+      "grad_norm": 0.009149067103862762,
+      "learning_rate": 0.0024306569343065694,
+      "loss": 1.1882,
+      "step": 112
+    },
+    {
+      "epoch": 0.24835164835164836,
+      "grad_norm": 0.0036444987636059523,
+      "learning_rate": 0.0024525547445255476,
+      "loss": 0.9138,
+      "step": 113
+    },
+    {
+      "epoch": 0.25054945054945055,
+      "grad_norm": 0.008033153600990772,
+      "learning_rate": 0.0024744525547445257,
+      "loss": 1.1289,
+      "step": 114
+    },
+    {
+      "epoch": 0.25274725274725274,
+      "grad_norm": 0.004118746612221003,
+      "learning_rate": 0.0024963503649635035,
+      "loss": 1.0117,
+      "step": 115
+    },
+    {
+      "epoch": 0.2549450549450549,
+      "grad_norm": 0.0040017166174948215,
+      "learning_rate": 0.0025182481751824816,
+      "loss": 0.8518,
+      "step": 116
+    },
+    {
+      "epoch": 0.2571428571428571,
+      "grad_norm": 0.0022865436039865017,
+      "learning_rate": 0.00254014598540146,
+      "loss": 0.9437,
+      "step": 117
+    },
+    {
+      "epoch": 0.25934065934065936,
+      "grad_norm": 0.013020399026572704,
+      "learning_rate": 0.002562043795620438,
+      "loss": 0.9184,
+      "step": 118
+    },
+    {
+      "epoch": 0.26153846153846155,
+      "grad_norm": 0.004709787666797638,
+      "learning_rate": 0.002583941605839416,
+      "loss": 0.9183,
+      "step": 119
+    },
+    {
+      "epoch": 0.26373626373626374,
+      "grad_norm": 0.005114705301821232,
+      "learning_rate": 0.0026058394160583943,
+      "loss": 0.9598,
+      "step": 120
+    },
+    {
+      "epoch": 0.26593406593406593,
+      "grad_norm": 0.003825002582743764,
+      "learning_rate": 0.0026277372262773725,
+      "loss": 0.9757,
+      "step": 121
+    },
+    {
+      "epoch": 0.2681318681318681,
+      "grad_norm": 0.003121759044006467,
+      "learning_rate": 0.0026496350364963507,
+      "loss": 0.9687,
+      "step": 122
+    },
+    {
+      "epoch": 0.2703296703296703,
+      "grad_norm": 0.006017250008881092,
+      "learning_rate": 0.0026715328467153284,
+      "loss": 0.9019,
+      "step": 123
+    },
+    {
+      "epoch": 0.2725274725274725,
+      "grad_norm": 0.00733914552256465,
+      "learning_rate": 0.0026934306569343066,
+      "loss": 0.9882,
+      "step": 124
+    },
+    {
+      "epoch": 0.27472527472527475,
+      "grad_norm": 0.004201827570796013,
+      "learning_rate": 0.0027153284671532847,
+      "loss": 1.0573,
+      "step": 125
+    },
+    {
+      "epoch": 0.27692307692307694,
+      "grad_norm": 0.008328534662723541,
+      "learning_rate": 0.002737226277372263,
+      "loss": 1.0574,
+      "step": 126
+    },
+    {
+      "epoch": 0.27912087912087913,
+      "grad_norm": 0.006531738676130772,
+      "learning_rate": 0.002759124087591241,
+      "loss": 0.9154,
+      "step": 127
+    },
+    {
+      "epoch": 0.2813186813186813,
+      "grad_norm": 0.00573331443592906,
+      "learning_rate": 0.002781021897810219,
+      "loss": 1.2329,
+      "step": 128
+    },
+    {
+      "epoch": 0.2835164835164835,
+      "grad_norm": 0.004643768072128296,
+      "learning_rate": 0.002802919708029197,
+      "loss": 0.8812,
+      "step": 129
+    },
+    {
+      "epoch": 0.2857142857142857,
+      "grad_norm": 0.004559768363833427,
+      "learning_rate": 0.002824817518248175,
+      "loss": 0.8749,
+      "step": 130
+    },
+    {
+      "epoch": 0.2879120879120879,
+      "grad_norm": 0.006374021526426077,
+      "learning_rate": 0.0028467153284671533,
+      "loss": 0.8898,
+      "step": 131
+    },
+    {
+      "epoch": 0.29010989010989013,
+      "grad_norm": 0.004299569875001907,
+      "learning_rate": 0.0028686131386861315,
+      "loss": 0.8569,
+      "step": 132
+    },
+    {
+      "epoch": 0.2923076923076923,
+      "grad_norm": 0.0038534703198820353,
+      "learning_rate": 0.0028905109489051097,
+      "loss": 1.0046,
+      "step": 133
+    },
+    {
+      "epoch": 0.2945054945054945,
+      "grad_norm": 0.003614328568801284,
+      "learning_rate": 0.002912408759124088,
+      "loss": 0.8193,
+      "step": 134
+    },
+    {
+      "epoch": 0.2967032967032967,
+      "grad_norm": 0.005756146740168333,
+      "learning_rate": 0.002934306569343066,
+      "loss": 1.1632,
+      "step": 135
+    },
+    {
+      "epoch": 0.2989010989010989,
+      "grad_norm": 0.003812008537352085,
+      "learning_rate": 0.0029562043795620437,
+      "loss": 1.1205,
+      "step": 136
+    },
+    {
+      "epoch": 0.3010989010989011,
+      "grad_norm": 0.004701883066445589,
+      "learning_rate": 0.002978102189781022,
+      "loss": 1.2408,
+      "step": 137
+    },
+    {
+      "epoch": 0.3032967032967033,
+      "grad_norm": 0.009417601861059666,
+      "learning_rate": 0.003,
+      "loss": 0.9867,
+      "step": 138
+    },
+    {
+      "epoch": 0.3054945054945055,
+      "grad_norm": 0.006507673300802708,
+      "learning_rate": 0.002997557003257329,
+      "loss": 0.7693,
+      "step": 139
+    },
+    {
+      "epoch": 0.3076923076923077,
+      "grad_norm": 0.00826818123459816,
+      "learning_rate": 0.002995114006514658,
+      "loss": 1.0,
+      "step": 140
+    },
+    {
+      "epoch": 0.3098901098901099,
+      "grad_norm": 0.006646919064223766,
+      "learning_rate": 0.002992671009771987,
+      "loss": 0.9817,
+      "step": 141
+    },
+    {
+      "epoch": 0.3120879120879121,
+      "grad_norm": 0.002918950514867902,
+      "learning_rate": 0.002990228013029316,
+      "loss": 0.9992,
+      "step": 142
+    },
+    {
+      "epoch": 0.3142857142857143,
+      "grad_norm": 0.006537210661917925,
+      "learning_rate": 0.002987785016286645,
+      "loss": 1.0357,
+      "step": 143
+    },
+    {
+      "epoch": 0.31648351648351647,
+      "grad_norm": 0.0029910150915384293,
+      "learning_rate": 0.002985342019543974,
+      "loss": 0.8484,
+      "step": 144
+    },
+    {
+      "epoch": 0.31868131868131866,
+      "grad_norm": 0.006709406152367592,
+      "learning_rate": 0.002982899022801303,
+      "loss": 0.9146,
+      "step": 145
+    },
+    {
+      "epoch": 0.3208791208791209,
+      "grad_norm": 0.005750632844865322,
+      "learning_rate": 0.002980456026058632,
+      "loss": 1.0595,
+      "step": 146
+    },
+    {
+      "epoch": 0.3230769230769231,
+      "grad_norm": 0.007539853919297457,
+      "learning_rate": 0.002978013029315961,
+      "loss": 1.0239,
+      "step": 147
+    },
+    {
+      "epoch": 0.3252747252747253,
+      "grad_norm": 0.008102650754153728,
+      "learning_rate": 0.00297557003257329,
+      "loss": 1.2216,
+      "step": 148
+    },
+    {
+      "epoch": 0.3274725274725275,
+      "grad_norm": 0.009737853892147541,
+      "learning_rate": 0.0029731270358306187,
+      "loss": 1.0042,
+      "step": 149
+    },
+    {
+      "epoch": 0.32967032967032966,
+      "grad_norm": 0.006044755689799786,
+      "learning_rate": 0.002970684039087948,
+      "loss": 1.0128,
+      "step": 150
+    },
+    {
+      "epoch": 0.33186813186813185,
+      "grad_norm": 0.009952268563210964,
+      "learning_rate": 0.002968241042345277,
+      "loss": 0.8561,
+      "step": 151
+    },
+    {
+      "epoch": 0.33406593406593404,
+      "grad_norm": 0.0035665163304656744,
+      "learning_rate": 0.002965798045602606,
+      "loss": 1.0493,
+      "step": 152
+    },
+    {
+      "epoch": 0.3362637362637363,
+      "grad_norm": 0.004359772428870201,
+      "learning_rate": 0.0029633550488599347,
+      "loss": 0.935,
+      "step": 153
+    },
+    {
+      "epoch": 0.3384615384615385,
+      "grad_norm": 0.007619280833750963,
+      "learning_rate": 0.002960912052117264,
+      "loss": 0.9689,
+      "step": 154
+    },
+    {
+      "epoch": 0.34065934065934067,
+      "grad_norm": 0.0057828971184790134,
+      "learning_rate": 0.002958469055374593,
+      "loss": 1.0908,
+      "step": 155
+    },
+    {
+      "epoch": 0.34285714285714286,
+      "grad_norm": 0.005319708492606878,
+      "learning_rate": 0.002956026058631922,
+      "loss": 0.9256,
+      "step": 156
+    },
+    {
+      "epoch": 0.34505494505494505,
+      "grad_norm": 0.004901966080069542,
+      "learning_rate": 0.0029535830618892507,
+      "loss": 0.8818,
+      "step": 157
+    },
+    {
+      "epoch": 0.34725274725274724,
+      "grad_norm": 0.004528634250164032,
+      "learning_rate": 0.0029511400651465796,
+      "loss": 1.2328,
+      "step": 158
+    },
+    {
+      "epoch": 0.34945054945054943,
+      "grad_norm": 0.00630349013954401,
+      "learning_rate": 0.002948697068403909,
+      "loss": 0.8034,
+      "step": 159
+    },
+    {
+      "epoch": 0.3516483516483517,
+      "grad_norm": 0.005868111737072468,
+      "learning_rate": 0.002946254071661238,
+      "loss": 1.1222,
+      "step": 160
+    },
+    {
+      "epoch": 0.35384615384615387,
+      "grad_norm": 0.006086320616304874,
+      "learning_rate": 0.0029438110749185667,
+      "loss": 1.1235,
+      "step": 161
+    },
+    {
+      "epoch": 0.35604395604395606,
+      "grad_norm": 0.006224587094038725,
+      "learning_rate": 0.0029413680781758956,
+      "loss": 0.9842,
+      "step": 162
+    },
+    {
+      "epoch": 0.35824175824175825,
+      "grad_norm": 0.005941646173596382,
+      "learning_rate": 0.002938925081433225,
+      "loss": 0.8085,
+      "step": 163
+    },
+    {
+      "epoch": 0.36043956043956044,
+      "grad_norm": 0.005233312491327524,
+      "learning_rate": 0.002936482084690554,
+      "loss": 0.9295,
+      "step": 164
+    },
+    {
+      "epoch": 0.3626373626373626,
+      "grad_norm": 0.007866628468036652,
+      "learning_rate": 0.0029340390879478827,
+      "loss": 0.9318,
+      "step": 165
+    },
+    {
+      "epoch": 0.3648351648351648,
+      "grad_norm": 0.006403940264135599,
+      "learning_rate": 0.0029315960912052116,
+      "loss": 0.9713,
+      "step": 166
+    },
+    {
+      "epoch": 0.367032967032967,
+      "grad_norm": 0.007658285088837147,
+      "learning_rate": 0.0029291530944625405,
+      "loss": 0.8402,
+      "step": 167
+    },
+    {
+      "epoch": 0.36923076923076925,
+      "grad_norm": 0.007293926551938057,
+      "learning_rate": 0.00292671009771987,
+      "loss": 1.0002,
+      "step": 168
+    },
+    {
+      "epoch": 0.37142857142857144,
+      "grad_norm": 0.005089641083031893,
+      "learning_rate": 0.0029242671009771988,
+      "loss": 0.8762,
+      "step": 169
+    },
+    {
+      "epoch": 0.37362637362637363,
+      "grad_norm": 0.006225347053259611,
+      "learning_rate": 0.0029218241042345276,
+      "loss": 0.9706,
+      "step": 170
+    },
+    {
+      "epoch": 0.3758241758241758,
+      "grad_norm": 0.006968878675252199,
+      "learning_rate": 0.0029193811074918565,
+      "loss": 0.9262,
+      "step": 171
+    },
+    {
+      "epoch": 0.378021978021978,
+      "grad_norm": 0.005770348012447357,
+      "learning_rate": 0.002916938110749186,
+      "loss": 1.0072,
+      "step": 172
+    },
+    {
+      "epoch": 0.3802197802197802,
+      "grad_norm": 0.00958232395350933,
+      "learning_rate": 0.0029144951140065148,
+      "loss": 0.8597,
+      "step": 173
+    },
+    {
+      "epoch": 0.3824175824175824,
+      "grad_norm": 0.0052667479030787945,
+      "learning_rate": 0.0029120521172638436,
+      "loss": 0.9212,
+      "step": 174
+    },
+    {
+      "epoch": 0.38461538461538464,
+      "grad_norm": 0.005888691637665033,
+      "learning_rate": 0.0029096091205211725,
+      "loss": 0.8979,
+      "step": 175
+    },
+    {
+      "epoch": 0.3868131868131868,
+      "grad_norm": 0.00905114971101284,
+      "learning_rate": 0.002907166123778502,
+      "loss": 1.164,
+      "step": 176
+    },
+    {
+      "epoch": 0.389010989010989,
+      "grad_norm": 0.007187289651483297,
+      "learning_rate": 0.0029047231270358308,
+      "loss": 0.8966,
+      "step": 177
+    },
+    {
+      "epoch": 0.3912087912087912,
+      "grad_norm": 0.006833909545093775,
+      "learning_rate": 0.0029022801302931596,
+      "loss": 0.9517,
+      "step": 178
+    },
+    {
+      "epoch": 0.3934065934065934,
+      "grad_norm": 0.006230995524674654,
+      "learning_rate": 0.0028998371335504885,
+      "loss": 0.9002,
+      "step": 179
+    },
+    {
+      "epoch": 0.3956043956043956,
+      "grad_norm": 0.0069520678371191025,
+      "learning_rate": 0.0028973941368078174,
+      "loss": 0.8897,
+      "step": 180
+    },
+    {
+      "epoch": 0.3978021978021978,
+      "grad_norm": 0.0065953293815255165,
+      "learning_rate": 0.0028949511400651468,
+      "loss": 0.9838,
+      "step": 181
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.006523432210087776,
+      "learning_rate": 0.0028925081433224756,
+      "loss": 0.9521,
+      "step": 182
+    },
+    {
+      "epoch": 0.4021978021978022,
+      "grad_norm": 0.0067771561443805695,
+      "learning_rate": 0.0028900651465798045,
+      "loss": 0.945,
+      "step": 183
+    },
+    {
+      "epoch": 0.4043956043956044,
+      "grad_norm": 0.008103212341666222,
+      "learning_rate": 0.0028876221498371334,
+      "loss": 1.0373,
+      "step": 184
+    },
+    {
+      "epoch": 0.4065934065934066,
+      "grad_norm": 0.00836127158254385,
+      "learning_rate": 0.0028851791530944628,
+      "loss": 0.9119,
+      "step": 185
+    },
+    {
+      "epoch": 0.4087912087912088,
+      "grad_norm": 0.019597329199314117,
+      "learning_rate": 0.0028827361563517916,
+      "loss": 0.8359,
+      "step": 186
+    },
+    {
+      "epoch": 0.41098901098901097,
+      "grad_norm": 0.006631451193243265,
+      "learning_rate": 0.0028802931596091205,
+      "loss": 1.0172,
+      "step": 187
+    },
+    {
+      "epoch": 0.41318681318681316,
+      "grad_norm": 0.009372205473482609,
+      "learning_rate": 0.0028778501628664494,
+      "loss": 1.007,
+      "step": 188
+    },
+    {
+      "epoch": 0.4153846153846154,
+      "grad_norm": 0.0077335285022854805,
+      "learning_rate": 0.0028754071661237783,
+      "loss": 1.007,
+      "step": 189
+    },
+    {
+      "epoch": 0.4175824175824176,
+      "grad_norm": 0.010944009758532047,
+      "learning_rate": 0.0028729641693811077,
+      "loss": 1.0769,
+      "step": 190
+    },
+    {
+      "epoch": 0.4197802197802198,
+      "grad_norm": 0.00904717668890953,
+      "learning_rate": 0.0028705211726384365,
+      "loss": 0.9313,
+      "step": 191
+    },
+    {
+      "epoch": 0.421978021978022,
+      "grad_norm": 0.010331280529499054,
+      "learning_rate": 0.0028680781758957654,
+      "loss": 1.0601,
+      "step": 192
+    },
+    {
+      "epoch": 0.42417582417582417,
+      "grad_norm": 0.006422454956918955,
+      "learning_rate": 0.0028656351791530943,
+      "loss": 0.9459,
+      "step": 193
+    },
+    {
+      "epoch": 0.42637362637362636,
+      "grad_norm": 0.006438535638153553,
+      "learning_rate": 0.0028631921824104237,
+      "loss": 0.9428,
+      "step": 194
+    },
+    {
+      "epoch": 0.42857142857142855,
+      "grad_norm": 0.006776008289307356,
+      "learning_rate": 0.0028607491856677525,
+      "loss": 0.9184,
+      "step": 195
+    },
+    {
+      "epoch": 0.4307692307692308,
+      "grad_norm": 0.009496882557868958,
+      "learning_rate": 0.0028583061889250814,
+      "loss": 0.9601,
+      "step": 196
+    },
+    {
+      "epoch": 0.432967032967033,
+      "grad_norm": 0.009344259276986122,
+      "learning_rate": 0.0028558631921824103,
+      "loss": 0.8633,
+      "step": 197
+    },
+    {
+      "epoch": 0.4351648351648352,
+      "grad_norm": 0.005349337123334408,
+      "learning_rate": 0.0028534201954397392,
+      "loss": 0.8971,
+      "step": 198
+    },
+    {
+      "epoch": 0.43736263736263736,
+      "grad_norm": 0.00725796353071928,
+      "learning_rate": 0.0028509771986970685,
+      "loss": 0.9274,
+      "step": 199
+    },
+    {
+      "epoch": 0.43956043956043955,
+      "grad_norm": 0.007469319272786379,
+      "learning_rate": 0.0028485342019543974,
+      "loss": 0.8673,
+      "step": 200
+    },
+    {
+      "epoch": 0.44175824175824174,
+      "grad_norm": 0.004618263803422451,
+      "learning_rate": 0.0028460912052117263,
+      "loss": 0.8125,
+      "step": 201
+    },
+    {
+      "epoch": 0.44395604395604393,
+      "grad_norm": 0.006519709713757038,
+      "learning_rate": 0.0028436482084690552,
+      "loss": 0.9718,
+      "step": 202
+    },
+    {
+      "epoch": 0.4461538461538462,
+      "grad_norm": 0.007761015556752682,
+      "learning_rate": 0.0028412052117263845,
+      "loss": 1.0727,
+      "step": 203
+    },
+    {
+      "epoch": 0.44835164835164837,
+      "grad_norm": 0.005366064142435789,
+      "learning_rate": 0.0028387622149837134,
+      "loss": 0.7338,
+      "step": 204
+    },
+    {
+      "epoch": 0.45054945054945056,
+      "grad_norm": 0.007449703756719828,
+      "learning_rate": 0.0028363192182410423,
+      "loss": 0.8927,
+      "step": 205
+    },
+    {
+      "epoch": 0.45274725274725275,
+      "grad_norm": 0.004439018201082945,
+      "learning_rate": 0.0028338762214983712,
+      "loss": 0.934,
+      "step": 206
+    },
+    {
+      "epoch": 0.45494505494505494,
+      "grad_norm": 0.0062678805552423,
+      "learning_rate": 0.0028314332247557,
+      "loss": 0.8323,
+      "step": 207
+    },
+    {
+      "epoch": 0.45714285714285713,
+      "grad_norm": 0.007749352138489485,
+      "learning_rate": 0.0028289902280130294,
+      "loss": 0.9213,
+      "step": 208
+    },
+    {
+      "epoch": 0.4593406593406593,
+      "grad_norm": 0.009455111809074879,
+      "learning_rate": 0.0028265472312703583,
+      "loss": 0.8485,
+      "step": 209
+    },
+    {
+      "epoch": 0.46153846153846156,
+      "grad_norm": 0.008707991801202297,
+      "learning_rate": 0.0028241042345276872,
+      "loss": 1.0288,
+      "step": 210
+    },
+    {
+      "epoch": 0.46373626373626375,
+      "grad_norm": 0.005036712158471346,
+      "learning_rate": 0.002821661237785016,
+      "loss": 0.8297,
+      "step": 211
+    },
+    {
+      "epoch": 0.46593406593406594,
+      "grad_norm": 0.005751885939389467,
+      "learning_rate": 0.0028192182410423454,
+      "loss": 0.7929,
+      "step": 212
+    },
+    {
+      "epoch": 0.46813186813186813,
+      "grad_norm": 0.004680875223129988,
+      "learning_rate": 0.0028167752442996743,
+      "loss": 0.9667,
+      "step": 213
+    },
+    {
+      "epoch": 0.4703296703296703,
+      "grad_norm": 0.005063106305897236,
+      "learning_rate": 0.0028143322475570032,
+      "loss": 0.8135,
+      "step": 214
+    },
+    {
+      "epoch": 0.4725274725274725,
+      "grad_norm": 0.005444399081170559,
+      "learning_rate": 0.002811889250814332,
+      "loss": 0.7836,
+      "step": 215
+    },
+    {
+      "epoch": 0.4747252747252747,
+      "grad_norm": 0.005630222614854574,
+      "learning_rate": 0.002809446254071661,
+      "loss": 0.975,
+      "step": 216
+    },
+    {
+      "epoch": 0.47692307692307695,
+      "grad_norm": 0.01288183405995369,
+      "learning_rate": 0.0028070032573289903,
+      "loss": 1.054,
+      "step": 217
+    },
+    {
+      "epoch": 0.47912087912087914,
+      "grad_norm": 0.006149762775748968,
+      "learning_rate": 0.0028045602605863192,
+      "loss": 0.8891,
+      "step": 218
+    },
+    {
+      "epoch": 0.48131868131868133,
+      "grad_norm": 0.005106229800730944,
+      "learning_rate": 0.002802117263843648,
+      "loss": 0.8609,
+      "step": 219
+    },
+    {
+      "epoch": 0.4835164835164835,
+      "grad_norm": 0.006657483521848917,
+      "learning_rate": 0.002799674267100977,
+      "loss": 0.769,
+      "step": 220
+    },
+    {
+      "epoch": 0.4857142857142857,
+      "grad_norm": 0.007483678869903088,
+      "learning_rate": 0.0027972312703583063,
+      "loss": 0.8018,
+      "step": 221
+    },
+    {
+      "epoch": 0.4879120879120879,
+      "grad_norm": 0.0057559930719435215,
+      "learning_rate": 0.0027947882736156352,
+      "loss": 0.9391,
+      "step": 222
+    },
+    {
+      "epoch": 0.4901098901098901,
+      "grad_norm": 0.006948585622012615,
+      "learning_rate": 0.002792345276872964,
+      "loss": 0.7839,
+      "step": 223
+    },
+    {
+      "epoch": 0.49230769230769234,
+      "grad_norm": 0.00421064393594861,
+      "learning_rate": 0.002789902280130293,
+      "loss": 0.7959,
+      "step": 224
+    },
+    {
+      "epoch": 0.4945054945054945,
+      "grad_norm": 0.009155537001788616,
+      "learning_rate": 0.002787459283387622,
+      "loss": 1.1663,
+      "step": 225
+    },
+    {
+      "epoch": 0.4967032967032967,
+      "grad_norm": 0.007379762828350067,
+      "learning_rate": 0.0027850162866449512,
+      "loss": 0.7917,
+      "step": 226
+    },
+    {
+      "epoch": 0.4989010989010989,
+      "grad_norm": 0.004164863843470812,
+      "learning_rate": 0.00278257328990228,
+      "loss": 0.7445,
+      "step": 227
+    },
+    {
+      "epoch": 0.5010989010989011,
+      "grad_norm": 0.00907017569988966,
+      "learning_rate": 0.002780130293159609,
+      "loss": 0.7838,
+      "step": 228
+    },
+    {
+      "epoch": 0.5032967032967033,
+      "grad_norm": 0.00643718708306551,
+      "learning_rate": 0.002777687296416938,
+      "loss": 0.8832,
+      "step": 229
+    },
+    {
+      "epoch": 0.5054945054945055,
+      "grad_norm": 0.010059970431029797,
+      "learning_rate": 0.0027752442996742672,
+      "loss": 0.9675,
+      "step": 230
+    },
+    {
+      "epoch": 0.5076923076923077,
+      "grad_norm": 0.02143646590411663,
+      "learning_rate": 0.002772801302931596,
+      "loss": 0.9997,
+      "step": 231
+    },
+    {
+      "epoch": 0.5098901098901099,
+      "grad_norm": 0.009404446929693222,
+      "learning_rate": 0.002770358306188925,
+      "loss": 0.8348,
+      "step": 232
+    },
+    {
+      "epoch": 0.512087912087912,
+      "grad_norm": 0.012918700464069843,
+      "learning_rate": 0.002767915309446254,
+      "loss": 1.1064,
+      "step": 233
+    },
+    {
+      "epoch": 0.5142857142857142,
+      "grad_norm": 0.004538615234196186,
+      "learning_rate": 0.002765472312703583,
+      "loss": 0.8897,
+      "step": 234
+    },
+    {
+      "epoch": 0.5164835164835165,
+      "grad_norm": 0.011407903395593166,
+      "learning_rate": 0.002763029315960912,
+      "loss": 1.0489,
+      "step": 235
+    },
+    {
+      "epoch": 0.5186813186813187,
+      "grad_norm": 0.005104811862111092,
+      "learning_rate": 0.002760586319218241,
+      "loss": 0.8202,
+      "step": 236
+    },
+    {
+      "epoch": 0.5208791208791209,
+      "grad_norm": 0.007670343387871981,
+      "learning_rate": 0.00275814332247557,
+      "loss": 0.7742,
+      "step": 237
+    },
+    {
+      "epoch": 0.5230769230769231,
+      "grad_norm": 0.005203275475651026,
+      "learning_rate": 0.002755700325732899,
+      "loss": 0.8876,
+      "step": 238
+    },
+    {
+      "epoch": 0.5252747252747253,
+      "grad_norm": 0.007757449056953192,
+      "learning_rate": 0.002753257328990228,
+      "loss": 0.7809,
+      "step": 239
+    },
+    {
+      "epoch": 0.5274725274725275,
+      "grad_norm": 0.007763593923300505,
+      "learning_rate": 0.002750814332247557,
+      "loss": 0.9395,
+      "step": 240
+    },
+    {
+      "epoch": 0.5296703296703297,
+      "grad_norm": 0.00601625069975853,
+      "learning_rate": 0.002748371335504886,
+      "loss": 0.7665,
+      "step": 241
+    },
+    {
+      "epoch": 0.5318681318681319,
+      "grad_norm": 0.005586160346865654,
+      "learning_rate": 0.002745928338762215,
+      "loss": 0.9336,
+      "step": 242
+    },
+    {
+      "epoch": 0.5340659340659341,
+      "grad_norm": 0.009924220852553844,
+      "learning_rate": 0.002743485342019544,
+      "loss": 0.8532,
+      "step": 243
+    },
+    {
+      "epoch": 0.5362637362637362,
+      "grad_norm": 0.007583543658256531,
+      "learning_rate": 0.002741042345276873,
+      "loss": 0.968,
+      "step": 244
+    },
+    {
+      "epoch": 0.5384615384615384,
+      "grad_norm": 0.006119458936154842,
+      "learning_rate": 0.002738599348534202,
+      "loss": 0.994,
+      "step": 245
+    },
+    {
+      "epoch": 0.5406593406593406,
+      "grad_norm": 0.005893225781619549,
+      "learning_rate": 0.002736156351791531,
+      "loss": 0.9101,
+      "step": 246
+    },
+    {
+      "epoch": 0.5428571428571428,
+      "grad_norm": 0.01102379523217678,
+      "learning_rate": 0.0027337133550488597,
+      "loss": 1.2117,
+      "step": 247
+    },
+    {
+      "epoch": 0.545054945054945,
+      "grad_norm": 0.004842462483793497,
+      "learning_rate": 0.002731270358306189,
+      "loss": 0.8792,
+      "step": 248
+    },
+    {
+      "epoch": 0.5472527472527473,
+      "grad_norm": 0.004212184343487024,
+      "learning_rate": 0.002728827361563518,
+      "loss": 0.7134,
+      "step": 249
+    },
+    {
+      "epoch": 0.5494505494505495,
+      "grad_norm": 0.00784846767783165,
+      "learning_rate": 0.002726384364820847,
+      "loss": 0.9464,
+      "step": 250
+    },
+    {
+      "epoch": 0.5516483516483517,
+      "grad_norm": 0.0067728846333920956,
+      "learning_rate": 0.0027239413680781757,
+      "loss": 0.9525,
+      "step": 251
+    },
+    {
+      "epoch": 0.5538461538461539,
+      "grad_norm": 0.005015241913497448,
+      "learning_rate": 0.002721498371335505,
+      "loss": 0.7694,
+      "step": 252
+    },
+    {
+      "epoch": 0.5560439560439561,
+      "grad_norm": 0.004611070267856121,
+      "learning_rate": 0.002719055374592834,
+      "loss": 0.8936,
+      "step": 253
+    },
+    {
+      "epoch": 0.5582417582417583,
+      "grad_norm": 0.007814732380211353,
+      "learning_rate": 0.002716612377850163,
+      "loss": 1.0438,
+      "step": 254
+    },
+    {
+      "epoch": 0.5604395604395604,
+      "grad_norm": 0.007011615205556154,
+      "learning_rate": 0.0027141693811074917,
+      "loss": 0.8333,
+      "step": 255
+    },
+    {
+      "epoch": 0.5626373626373626,
+      "grad_norm": 0.0050247847102582455,
+      "learning_rate": 0.0027117263843648206,
+      "loss": 0.8197,
+      "step": 256
+    },
+    {
+      "epoch": 0.5648351648351648,
+      "grad_norm": 0.006660814397037029,
+      "learning_rate": 0.00270928338762215,
+      "loss": 0.8253,
+      "step": 257
+    },
+    {
+      "epoch": 0.567032967032967,
+      "grad_norm": 0.0035946108400821686,
+      "learning_rate": 0.002706840390879479,
+      "loss": 0.9736,
+      "step": 258
+    },
+    {
+      "epoch": 0.5692307692307692,
+      "grad_norm": 0.0059481472708284855,
+      "learning_rate": 0.0027043973941368077,
+      "loss": 0.8544,
+      "step": 259
+    },
+    {
+      "epoch": 0.5714285714285714,
+      "grad_norm": 0.009644001722335815,
+      "learning_rate": 0.0027019543973941366,
+      "loss": 0.7548,
+      "step": 260
+    },
+    {
+      "epoch": 0.5736263736263736,
+      "grad_norm": 0.006974764633923769,
+      "learning_rate": 0.002699511400651466,
+      "loss": 0.8788,
+      "step": 261
+    },
+    {
+      "epoch": 0.5758241758241758,
+      "grad_norm": 0.0057715862058103085,
+      "learning_rate": 0.002697068403908795,
+      "loss": 0.821,
+      "step": 262
+    },
+    {
+      "epoch": 0.578021978021978,
+      "grad_norm": 0.008595791645348072,
+      "learning_rate": 0.0026946254071661237,
+      "loss": 0.8372,
+      "step": 263
+    },
+    {
+      "epoch": 0.5802197802197803,
+      "grad_norm": 0.009786032140254974,
+      "learning_rate": 0.0026921824104234526,
+      "loss": 0.849,
+      "step": 264
+    },
+    {
+      "epoch": 0.5824175824175825,
+      "grad_norm": 0.004562468733638525,
+      "learning_rate": 0.0026897394136807815,
+      "loss": 0.8997,
+      "step": 265
+    },
+    {
+      "epoch": 0.5846153846153846,
+      "grad_norm": 0.004977593198418617,
+      "learning_rate": 0.002687296416938111,
+      "loss": 0.8427,
+      "step": 266
+    },
+    {
+      "epoch": 0.5868131868131868,
+      "grad_norm": 0.006152510177344084,
+      "learning_rate": 0.0026848534201954397,
+      "loss": 0.7852,
+      "step": 267
+    },
+    {
+      "epoch": 0.589010989010989,
+      "grad_norm": 0.005706351716071367,
+      "learning_rate": 0.0026824104234527686,
+      "loss": 0.9278,
+      "step": 268
+    },
+    {
+      "epoch": 0.5912087912087912,
+      "grad_norm": 0.004968094639480114,
+      "learning_rate": 0.0026799674267100975,
+      "loss": 0.7733,
+      "step": 269
+    },
+    {
+      "epoch": 0.5934065934065934,
+      "grad_norm": 0.00936887226998806,
+      "learning_rate": 0.002677524429967427,
+      "loss": 0.8623,
+      "step": 270
+    },
+    {
+      "epoch": 0.5956043956043956,
+      "grad_norm": 0.00931707676500082,
+      "learning_rate": 0.0026750814332247557,
+      "loss": 0.9613,
+      "step": 271
+    },
+    {
+      "epoch": 0.5978021978021978,
+      "grad_norm": 0.008495870046317577,
+      "learning_rate": 0.0026726384364820846,
+      "loss": 0.7598,
+      "step": 272
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.016891246661543846,
+      "learning_rate": 0.0026701954397394135,
+      "loss": 1.1428,
+      "step": 273
+    },
+    {
+      "epoch": 0.6021978021978022,
+      "grad_norm": 0.00998606439679861,
+      "learning_rate": 0.0026677524429967424,
+      "loss": 0.8623,
+      "step": 274
+    },
+    {
+      "epoch": 0.6043956043956044,
+      "grad_norm": 0.009116140194237232,
+      "learning_rate": 0.0026653094462540717,
+      "loss": 0.9657,
+      "step": 275
+    },
+    {
+      "epoch": 0.6065934065934065,
+      "grad_norm": 0.006984033156186342,
+      "learning_rate": 0.0026628664495114006,
+      "loss": 0.9785,
+      "step": 276
+    },
+    {
+      "epoch": 0.6087912087912087,
+      "grad_norm": 0.0060313087888062,
+      "learning_rate": 0.0026604234527687295,
+      "loss": 0.8321,
+      "step": 277
+    },
+    {
+      "epoch": 0.610989010989011,
+      "grad_norm": 0.004724503494799137,
+      "learning_rate": 0.0026579804560260584,
+      "loss": 0.659,
+      "step": 278
+    },
+    {
+      "epoch": 0.6131868131868132,
+      "grad_norm": 0.007574123330414295,
+      "learning_rate": 0.0026555374592833877,
+      "loss": 0.7948,
+      "step": 279
+    },
+    {
+      "epoch": 0.6153846153846154,
+      "grad_norm": 0.006810783874243498,
+      "learning_rate": 0.0026530944625407166,
+      "loss": 0.7441,
+      "step": 280
+    },
+    {
+      "epoch": 0.6175824175824176,
+      "grad_norm": 0.005541827995330095,
+      "learning_rate": 0.0026506514657980455,
+      "loss": 0.7101,
+      "step": 281
+    },
+    {
+      "epoch": 0.6197802197802198,
+      "grad_norm": 0.005206754896789789,
+      "learning_rate": 0.0026482084690553744,
+      "loss": 0.6641,
+      "step": 282
+    },
+    {
+      "epoch": 0.621978021978022,
+      "grad_norm": 0.005909618455916643,
+      "learning_rate": 0.0026457654723127033,
+      "loss": 0.9081,
+      "step": 283
+    },
+    {
+      "epoch": 0.6241758241758242,
+      "grad_norm": 0.006171086803078651,
+      "learning_rate": 0.0026433224755700326,
+      "loss": 0.8435,
+      "step": 284
+    },
+    {
+      "epoch": 0.6263736263736264,
+      "grad_norm": 0.0065986234694719315,
+      "learning_rate": 0.0026408794788273615,
+      "loss": 0.8528,
+      "step": 285
+    },
+    {
+      "epoch": 0.6285714285714286,
+      "grad_norm": 0.011476404033601284,
+      "learning_rate": 0.0026384364820846904,
+      "loss": 0.8913,
+      "step": 286
+    },
+    {
+      "epoch": 0.6307692307692307,
+      "grad_norm": 0.004490249324589968,
+      "learning_rate": 0.0026359934853420193,
+      "loss": 0.825,
+      "step": 287
+    },
+    {
+      "epoch": 0.6329670329670329,
+      "grad_norm": 0.005908684339374304,
+      "learning_rate": 0.0026335504885993486,
+      "loss": 0.9567,
+      "step": 288
+    },
+    {
+      "epoch": 0.6351648351648351,
+      "grad_norm": 0.005294302944093943,
+      "learning_rate": 0.0026311074918566775,
+      "loss": 0.8834,
+      "step": 289
+    },
+    {
+      "epoch": 0.6373626373626373,
+      "grad_norm": 0.0070920041762292385,
+      "learning_rate": 0.0026286644951140064,
+      "loss": 0.8737,
+      "step": 290
+    },
+    {
+      "epoch": 0.6395604395604395,
+      "grad_norm": 0.005321511533111334,
+      "learning_rate": 0.0026262214983713353,
+      "loss": 0.8357,
+      "step": 291
+    },
+    {
+      "epoch": 0.6417582417582418,
+      "grad_norm": 0.0073551912792027,
+      "learning_rate": 0.0026237785016286646,
+      "loss": 0.7596,
+      "step": 292
+    },
+    {
+      "epoch": 0.643956043956044,
+      "grad_norm": 0.006069947965443134,
+      "learning_rate": 0.0026213355048859935,
+      "loss": 0.7705,
+      "step": 293
+    },
+    {
+      "epoch": 0.6461538461538462,
+      "grad_norm": 0.05065399035811424,
+      "learning_rate": 0.002618892508143323,
+      "loss": 0.8105,
+      "step": 294
+    },
+    {
+      "epoch": 0.6483516483516484,
+      "grad_norm": 0.005198487546294928,
+      "learning_rate": 0.0026164495114006517,
+      "loss": 0.752,
+      "step": 295
+    },
+    {
+      "epoch": 0.6505494505494506,
+      "grad_norm": 0.0123857781291008,
+      "learning_rate": 0.0026140065146579806,
+      "loss": 0.9236,
+      "step": 296
+    },
+    {
+      "epoch": 0.6527472527472528,
+      "grad_norm": 0.004860189277678728,
+      "learning_rate": 0.0026115635179153095,
+      "loss": 0.8956,
+      "step": 297
+    },
+    {
+      "epoch": 0.654945054945055,
+      "grad_norm": 0.008202974684536457,
+      "learning_rate": 0.002609120521172639,
+      "loss": 0.9025,
+      "step": 298
+    },
+    {
+      "epoch": 0.6571428571428571,
+      "grad_norm": 0.006815968547016382,
+      "learning_rate": 0.0026066775244299677,
+      "loss": 0.9766,
+      "step": 299
+    },
+    {
+      "epoch": 0.6593406593406593,
+      "grad_norm": 0.007751207333058119,
+      "learning_rate": 0.0026042345276872966,
+      "loss": 0.9059,
+      "step": 300
+    },
+    {
+      "epoch": 0.6615384615384615,
+      "grad_norm": 0.006594427861273289,
+      "learning_rate": 0.0026017915309446255,
+      "loss": 0.8649,
+      "step": 301
+    },
+    {
+      "epoch": 0.6637362637362637,
+      "grad_norm": 0.007775988895446062,
+      "learning_rate": 0.0025993485342019544,
+      "loss": 0.977,
+      "step": 302
+    },
+    {
+      "epoch": 0.6659340659340659,
+      "grad_norm": 0.00522700184956193,
+      "learning_rate": 0.0025969055374592837,
+      "loss": 0.7261,
+      "step": 303
+    },
+    {
+      "epoch": 0.6681318681318681,
+      "grad_norm": 0.0054281787015497684,
+      "learning_rate": 0.0025944625407166126,
+      "loss": 0.7643,
+      "step": 304
+    },
+    {
+      "epoch": 0.6703296703296703,
+      "grad_norm": 0.006651235278695822,
+      "learning_rate": 0.0025920195439739415,
+      "loss": 0.7715,
+      "step": 305
+    },
+    {
+      "epoch": 0.6725274725274726,
+      "grad_norm": 0.007272882852703333,
+      "learning_rate": 0.0025895765472312704,
+      "loss": 0.9132,
+      "step": 306
+    },
+    {
+      "epoch": 0.6747252747252748,
+      "grad_norm": 0.006148648448288441,
+      "learning_rate": 0.0025871335504885997,
+      "loss": 0.8185,
+      "step": 307
+    },
+    {
+      "epoch": 0.676923076923077,
+      "grad_norm": 0.006991968955844641,
+      "learning_rate": 0.0025846905537459286,
+      "loss": 1.0655,
+      "step": 308
+    },
+    {
+      "epoch": 0.6791208791208792,
+      "grad_norm": 0.005360438488423824,
+      "learning_rate": 0.0025822475570032575,
+      "loss": 0.6664,
+      "step": 309
+    },
+    {
+      "epoch": 0.6813186813186813,
+      "grad_norm": 0.005835953634232283,
+      "learning_rate": 0.0025798045602605864,
+      "loss": 0.9402,
+      "step": 310
+    },
+    {
+      "epoch": 0.6835164835164835,
+      "grad_norm": 0.008970585651695728,
+      "learning_rate": 0.0025773615635179153,
+      "loss": 0.8704,
+      "step": 311
+    },
+    {
+      "epoch": 0.6857142857142857,
+      "grad_norm": 0.00933810044080019,
+      "learning_rate": 0.0025749185667752446,
+      "loss": 0.8974,
+      "step": 312
+    },
+    {
+      "epoch": 0.6879120879120879,
+      "grad_norm": 0.005828648805618286,
+      "learning_rate": 0.0025724755700325735,
+      "loss": 0.7602,
+      "step": 313
+    },
+    {
+      "epoch": 0.6901098901098901,
+      "grad_norm": 0.0052863433957099915,
+      "learning_rate": 0.0025700325732899024,
+      "loss": 0.8009,
+      "step": 314
+    },
+    {
+      "epoch": 0.6923076923076923,
+      "grad_norm": 0.009425065480172634,
+      "learning_rate": 0.0025675895765472313,
+      "loss": 0.7524,
+      "step": 315
+    },
+    {
+      "epoch": 0.6945054945054945,
+      "grad_norm": 0.009360330179333687,
+      "learning_rate": 0.0025651465798045606,
+      "loss": 0.7505,
+      "step": 316
+    },
+    {
+      "epoch": 0.6967032967032967,
+      "grad_norm": 0.017736751586198807,
+      "learning_rate": 0.0025627035830618895,
+      "loss": 1.0804,
+      "step": 317
+    },
+    {
+      "epoch": 0.6989010989010989,
+      "grad_norm": 0.007758346386253834,
+      "learning_rate": 0.0025602605863192184,
+      "loss": 0.8713,
+      "step": 318
+    },
+    {
+      "epoch": 0.701098901098901,
+      "grad_norm": 0.006658450700342655,
+      "learning_rate": 0.0025578175895765473,
+      "loss": 0.9631,
+      "step": 319
+    },
+    {
+      "epoch": 0.7032967032967034,
+      "grad_norm": 0.007880093529820442,
+      "learning_rate": 0.0025553745928338766,
+      "loss": 0.9746,
+      "step": 320
+    },
+    {
+      "epoch": 0.7054945054945055,
+      "grad_norm": 0.007918466813862324,
+      "learning_rate": 0.0025529315960912055,
+      "loss": 0.9373,
+      "step": 321
+    },
+    {
+      "epoch": 0.7076923076923077,
+      "grad_norm": 0.008261624723672867,
+      "learning_rate": 0.0025504885993485344,
+      "loss": 0.7928,
+      "step": 322
+    },
+    {
+      "epoch": 0.7098901098901099,
+      "grad_norm": 0.0057929521426558495,
+      "learning_rate": 0.0025480456026058633,
+      "loss": 0.824,
+      "step": 323
+    },
+    {
+      "epoch": 0.7120879120879121,
+      "grad_norm": 0.0049140132032334805,
+      "learning_rate": 0.002545602605863192,
+      "loss": 0.74,
+      "step": 324
+    },
+    {
+      "epoch": 0.7142857142857143,
+      "grad_norm": 0.00839124247431755,
+      "learning_rate": 0.0025431596091205215,
+      "loss": 0.9808,
+      "step": 325
+    },
+    {
+      "epoch": 0.7164835164835165,
+      "grad_norm": 0.005810212809592485,
+      "learning_rate": 0.0025407166123778504,
+      "loss": 0.8513,
+      "step": 326
+    },
+    {
+      "epoch": 0.7186813186813187,
+      "grad_norm": 0.007351922802627087,
+      "learning_rate": 0.0025382736156351793,
+      "loss": 0.9284,
+      "step": 327
+    },
+    {
+      "epoch": 0.7208791208791209,
+      "grad_norm": 0.007793235592544079,
+      "learning_rate": 0.002535830618892508,
+      "loss": 0.8551,
+      "step": 328
+    },
+    {
+      "epoch": 0.7230769230769231,
+      "grad_norm": 0.006734553724527359,
+      "learning_rate": 0.0025333876221498375,
+      "loss": 0.6325,
+      "step": 329
+    },
+    {
+      "epoch": 0.7252747252747253,
+      "grad_norm": 0.009309055283665657,
+      "learning_rate": 0.0025309446254071664,
+      "loss": 0.8767,
+      "step": 330
+    },
+    {
+      "epoch": 0.7274725274725274,
+      "grad_norm": 0.008501403033733368,
+      "learning_rate": 0.0025285016286644953,
+      "loss": 0.8046,
+      "step": 331
+    },
+    {
+      "epoch": 0.7296703296703296,
+      "grad_norm": 0.007142470683902502,
+      "learning_rate": 0.002526058631921824,
+      "loss": 0.6164,
+      "step": 332
+    },
+    {
+      "epoch": 0.7318681318681318,
+      "grad_norm": 0.005270811729133129,
+      "learning_rate": 0.002523615635179153,
+      "loss": 0.8914,
+      "step": 333
+    },
+    {
+      "epoch": 0.734065934065934,
+      "grad_norm": 0.005286007188260555,
+      "learning_rate": 0.0025211726384364824,
+      "loss": 0.9486,
+      "step": 334
+    },
+    {
+      "epoch": 0.7362637362637363,
+      "grad_norm": 0.006734045222401619,
+      "learning_rate": 0.0025187296416938113,
+      "loss": 0.7912,
+      "step": 335
+    },
+    {
+      "epoch": 0.7384615384615385,
+      "grad_norm": 0.004491882398724556,
+      "learning_rate": 0.00251628664495114,
+      "loss": 0.8662,
+      "step": 336
+    },
+    {
+      "epoch": 0.7406593406593407,
+      "grad_norm": 0.010574889369308949,
+      "learning_rate": 0.002513843648208469,
+      "loss": 0.8939,
+      "step": 337
+    },
+    {
+      "epoch": 0.7428571428571429,
+      "grad_norm": 0.01118052564561367,
+      "learning_rate": 0.0025114006514657984,
+      "loss": 0.8561,
+      "step": 338
+    },
+    {
+      "epoch": 0.7450549450549451,
+      "grad_norm": 0.010498358868062496,
+      "learning_rate": 0.0025089576547231273,
+      "loss": 0.6887,
+      "step": 339
+    },
+    {
+      "epoch": 0.7472527472527473,
+      "grad_norm": 0.008385674096643925,
+      "learning_rate": 0.002506514657980456,
+      "loss": 0.7705,
+      "step": 340
+    },
+    {
+      "epoch": 0.7494505494505495,
+      "grad_norm": 0.010413667187094688,
+      "learning_rate": 0.002504071661237785,
+      "loss": 0.7808,
+      "step": 341
+    },
+    {
+      "epoch": 0.7516483516483516,
+      "grad_norm": 0.007883529178798199,
+      "learning_rate": 0.002501628664495114,
+      "loss": 0.75,
+      "step": 342
+    },
+    {
+      "epoch": 0.7538461538461538,
+      "grad_norm": 0.00899143610149622,
+      "learning_rate": 0.0024991856677524433,
+      "loss": 0.7187,
+      "step": 343
+    },
+    {
+      "epoch": 0.756043956043956,
+      "grad_norm": 0.005088643170893192,
+      "learning_rate": 0.002496742671009772,
+      "loss": 0.793,
+      "step": 344
+    },
+    {
+      "epoch": 0.7582417582417582,
+      "grad_norm": 0.0029062123503535986,
+      "learning_rate": 0.002494299674267101,
+      "loss": 0.5868,
+      "step": 345
+    },
+    {
+      "epoch": 0.7604395604395604,
+      "grad_norm": 0.006681826431304216,
+      "learning_rate": 0.00249185667752443,
+      "loss": 0.6684,
+      "step": 346
+    },
+    {
+      "epoch": 0.7626373626373626,
+      "grad_norm": 0.010312552563846111,
+      "learning_rate": 0.0024894136807817593,
+      "loss": 0.8908,
+      "step": 347
+    },
+    {
+      "epoch": 0.7648351648351648,
+      "grad_norm": 0.011571112088859081,
+      "learning_rate": 0.002486970684039088,
+      "loss": 0.7846,
+      "step": 348
+    },
+    {
+      "epoch": 0.7670329670329671,
+      "grad_norm": 0.0058892760425806046,
+      "learning_rate": 0.002484527687296417,
+      "loss": 0.8244,
+      "step": 349
+    },
+    {
+      "epoch": 0.7692307692307693,
+      "grad_norm": 0.011332704685628414,
+      "learning_rate": 0.002482084690553746,
+      "loss": 0.8406,
+      "step": 350
+    },
+    {
+      "epoch": 0.7714285714285715,
+      "grad_norm": 0.005350037477910519,
+      "learning_rate": 0.002479641693811075,
+      "loss": 0.8625,
+      "step": 351
+    },
+    {
+      "epoch": 0.7736263736263737,
+      "grad_norm": 0.004495049361139536,
+      "learning_rate": 0.002477198697068404,
+      "loss": 0.6675,
+      "step": 352
+    },
+    {
+      "epoch": 0.7758241758241758,
+      "grad_norm": 0.005853444337844849,
+      "learning_rate": 0.002474755700325733,
+      "loss": 0.849,
+      "step": 353
+    },
+    {
+      "epoch": 0.778021978021978,
+      "grad_norm": 0.00924062542617321,
+      "learning_rate": 0.002472312703583062,
+      "loss": 0.8281,
+      "step": 354
+    },
+    {
+      "epoch": 0.7802197802197802,
+      "grad_norm": 0.008099173195660114,
+      "learning_rate": 0.002469869706840391,
+      "loss": 0.9685,
+      "step": 355
+    },
+    {
+      "epoch": 0.7824175824175824,
+      "grad_norm": 0.008459962904453278,
+      "learning_rate": 0.00246742671009772,
+      "loss": 0.892,
+      "step": 356
+    },
+    {
+      "epoch": 0.7846153846153846,
+      "grad_norm": 0.008643421344459057,
+      "learning_rate": 0.002464983713355049,
+      "loss": 0.6816,
+      "step": 357
+    },
+    {
+      "epoch": 0.7868131868131868,
+      "grad_norm": 0.021500252187252045,
+      "learning_rate": 0.002462540716612378,
+      "loss": 0.8956,
+      "step": 358
+    },
+    {
+      "epoch": 0.789010989010989,
+      "grad_norm": 0.008828614838421345,
+      "learning_rate": 0.002460097719869707,
+      "loss": 0.7828,
+      "step": 359
+    },
+    {
+      "epoch": 0.7912087912087912,
+      "grad_norm": 0.010325809009373188,
+      "learning_rate": 0.0024576547231270358,
+      "loss": 0.778,
+      "step": 360
+    },
+    {
+      "epoch": 0.7934065934065934,
+      "grad_norm": 0.005172448698431253,
+      "learning_rate": 0.002455211726384365,
+      "loss": 0.8197,
+      "step": 361
+    },
+    {
+      "epoch": 0.7956043956043956,
+      "grad_norm": 0.0072929454036056995,
+      "learning_rate": 0.002452768729641694,
+      "loss": 0.7481,
+      "step": 362
+    },
+    {
+      "epoch": 0.7978021978021979,
+      "grad_norm": 0.010980704799294472,
+      "learning_rate": 0.002450325732899023,
+      "loss": 1.0486,
+      "step": 363
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.007416799198836088,
+      "learning_rate": 0.0024478827361563518,
+      "loss": 0.6616,
+      "step": 364
+    },
+    {
+      "epoch": 0.8021978021978022,
+      "grad_norm": 0.016207987442612648,
+      "learning_rate": 0.002445439739413681,
+      "loss": 0.9203,
+      "step": 365
+    },
+    {
+      "epoch": 0.8043956043956044,
+      "grad_norm": 0.007297312840819359,
+      "learning_rate": 0.00244299674267101,
+      "loss": 0.9618,
+      "step": 366
+    },
+    {
+      "epoch": 0.8065934065934066,
+      "grad_norm": 0.007284694351255894,
+      "learning_rate": 0.002440553745928339,
+      "loss": 0.8593,
+      "step": 367
+    },
+    {
+      "epoch": 0.8087912087912088,
+      "grad_norm": 0.004910862073302269,
+      "learning_rate": 0.0024381107491856678,
+      "loss": 0.8115,
+      "step": 368
+    },
+    {
+      "epoch": 0.810989010989011,
+      "grad_norm": 0.007959703914821148,
+      "learning_rate": 0.0024356677524429967,
+      "loss": 0.7998,
+      "step": 369
+    },
+    {
+      "epoch": 0.8131868131868132,
+      "grad_norm": 0.006520700175315142,
+      "learning_rate": 0.002433224755700326,
+      "loss": 0.7909,
+      "step": 370
+    },
+    {
+      "epoch": 0.8153846153846154,
+      "grad_norm": 0.004056925885379314,
+      "learning_rate": 0.002430781758957655,
+      "loss": 0.9345,
+      "step": 371
+    },
+    {
+      "epoch": 0.8175824175824176,
+      "grad_norm": 0.00512728001922369,
+      "learning_rate": 0.0024283387622149838,
+      "loss": 0.8488,
+      "step": 372
+    },
+    {
+      "epoch": 0.8197802197802198,
+      "grad_norm": 0.008944818750023842,
+      "learning_rate": 0.0024258957654723127,
+      "loss": 0.7108,
+      "step": 373
+    },
+    {
+      "epoch": 0.8219780219780219,
+      "grad_norm": 0.006004352122545242,
+      "learning_rate": 0.002423452768729642,
+      "loss": 0.8041,
+      "step": 374
+    },
+    {
+      "epoch": 0.8241758241758241,
+      "grad_norm": 0.008614517748355865,
+      "learning_rate": 0.002421009771986971,
+      "loss": 0.7143,
+      "step": 375
+    },
+    {
+      "epoch": 0.8263736263736263,
+      "grad_norm": 0.017695611342787743,
+      "learning_rate": 0.0024185667752442998,
+      "loss": 1.008,
+      "step": 376
+    },
+    {
+      "epoch": 0.8285714285714286,
+      "grad_norm": 0.010052197612822056,
+      "learning_rate": 0.0024161237785016287,
+      "loss": 0.9878,
+      "step": 377
+    },
+    {
+      "epoch": 0.8307692307692308,
+      "grad_norm": 0.008998232893645763,
+      "learning_rate": 0.0024136807817589576,
+      "loss": 0.8495,
+      "step": 378
+    },
+    {
+      "epoch": 0.832967032967033,
+      "grad_norm": 0.00542048504576087,
+      "learning_rate": 0.002411237785016287,
+      "loss": 0.7757,
+      "step": 379
+    },
+    {
+      "epoch": 0.8351648351648352,
+      "grad_norm": 0.009006396867334843,
+      "learning_rate": 0.0024087947882736158,
+      "loss": 0.7513,
+      "step": 380
+    },
+    {
+      "epoch": 0.8373626373626374,
+      "grad_norm": 0.011888355948030949,
+      "learning_rate": 0.0024063517915309447,
+      "loss": 0.8903,
+      "step": 381
+    },
+    {
+      "epoch": 0.8395604395604396,
+      "grad_norm": 0.00523636769503355,
+      "learning_rate": 0.0024039087947882736,
+      "loss": 0.7318,
+      "step": 382
+    },
+    {
+      "epoch": 0.8417582417582418,
+      "grad_norm": 0.007836324162781239,
+      "learning_rate": 0.002401465798045603,
+      "loss": 0.8052,
+      "step": 383
+    },
+    {
+      "epoch": 0.843956043956044,
+      "grad_norm": 0.014790039509534836,
+      "learning_rate": 0.0023990228013029318,
+      "loss": 0.7785,
+      "step": 384
+    },
+    {
+      "epoch": 0.8461538461538461,
+      "grad_norm": 0.007122506387531757,
+      "learning_rate": 0.0023965798045602607,
+      "loss": 0.7836,
+      "step": 385
+    },
+    {
+      "epoch": 0.8483516483516483,
+      "grad_norm": 0.007114512380212545,
+      "learning_rate": 0.0023941368078175896,
+      "loss": 0.8652,
+      "step": 386
+    },
+    {
+      "epoch": 0.8505494505494505,
+      "grad_norm": 0.005843424703925848,
+      "learning_rate": 0.002391693811074919,
+      "loss": 1.1135,
+      "step": 387
+    },
+    {
+      "epoch": 0.8527472527472527,
+      "grad_norm": 0.010878733359277248,
+      "learning_rate": 0.0023892508143322478,
+      "loss": 0.7793,
+      "step": 388
+    },
+    {
+      "epoch": 0.8549450549450549,
+      "grad_norm": 0.0034110695123672485,
+      "learning_rate": 0.0023868078175895767,
+      "loss": 0.8883,
+      "step": 389
+    },
+    {
+      "epoch": 0.8571428571428571,
+      "grad_norm": 0.005664998199790716,
+      "learning_rate": 0.0023843648208469056,
+      "loss": 0.8166,
+      "step": 390
+    },
+    {
+      "epoch": 0.8593406593406593,
+      "grad_norm": 0.00942701380699873,
+      "learning_rate": 0.0023819218241042345,
+      "loss": 0.8787,
+      "step": 391
+    },
+    {
+      "epoch": 0.8615384615384616,
+      "grad_norm": 0.005407244432717562,
+      "learning_rate": 0.002379478827361564,
+      "loss": 0.8477,
+      "step": 392
+    },
+    {
+      "epoch": 0.8637362637362638,
+      "grad_norm": 0.005061489995568991,
+      "learning_rate": 0.0023770358306188927,
+      "loss": 0.7849,
+      "step": 393
+    },
+    {
+      "epoch": 0.865934065934066,
+      "grad_norm": 0.005626164376735687,
+      "learning_rate": 0.0023745928338762216,
+      "loss": 0.899,
+      "step": 394
+    },
+    {
+      "epoch": 0.8681318681318682,
+      "grad_norm": 0.005086739547550678,
+      "learning_rate": 0.0023721498371335505,
+      "loss": 0.8585,
+      "step": 395
+    },
+    {
+      "epoch": 0.8703296703296703,
+      "grad_norm": 0.006302012596279383,
+      "learning_rate": 0.00236970684039088,
+      "loss": 0.9078,
+      "step": 396
+    },
+    {
+      "epoch": 0.8725274725274725,
+      "grad_norm": 0.005908872000873089,
+      "learning_rate": 0.0023672638436482087,
+      "loss": 0.7307,
+      "step": 397
+    },
+    {
+      "epoch": 0.8747252747252747,
+      "grad_norm": 0.006952814292162657,
+      "learning_rate": 0.0023648208469055376,
+      "loss": 0.9338,
+      "step": 398
+    },
+    {
+      "epoch": 0.8769230769230769,
+      "grad_norm": 0.005616925191134214,
+      "learning_rate": 0.0023623778501628665,
+      "loss": 0.6998,
+      "step": 399
+    },
+    {
+      "epoch": 0.8791208791208791,
+      "grad_norm": 0.0057669165544211864,
+      "learning_rate": 0.0023599348534201954,
+      "loss": 0.8369,
+      "step": 400
+    },
+    {
+      "epoch": 0.8813186813186813,
+      "grad_norm": 0.004022832959890366,
+      "learning_rate": 0.0023574918566775247,
+      "loss": 0.8889,
+      "step": 401
+    },
+    {
+      "epoch": 0.8835164835164835,
+      "grad_norm": 0.004832746926695108,
+      "learning_rate": 0.0023550488599348536,
+      "loss": 0.837,
+      "step": 402
+    },
+    {
+      "epoch": 0.8857142857142857,
+      "grad_norm": 0.005809650756418705,
+      "learning_rate": 0.0023526058631921825,
+      "loss": 0.8055,
+      "step": 403
+    },
+    {
+      "epoch": 0.8879120879120879,
+      "grad_norm": 0.0059983632527291775,
+      "learning_rate": 0.0023501628664495114,
+      "loss": 0.7368,
+      "step": 404
+    },
+    {
+      "epoch": 0.8901098901098901,
+      "grad_norm": 0.0045838006772100925,
+      "learning_rate": 0.0023477198697068407,
+      "loss": 0.8492,
+      "step": 405
+    },
+    {
+      "epoch": 0.8923076923076924,
+      "grad_norm": 0.003913301043212414,
+      "learning_rate": 0.0023452768729641696,
+      "loss": 0.8012,
+      "step": 406
+    },
+    {
+      "epoch": 0.8945054945054945,
+      "grad_norm": 0.007567413616925478,
+      "learning_rate": 0.0023428338762214985,
+      "loss": 0.7469,
+      "step": 407
+    },
+    {
+      "epoch": 0.8967032967032967,
+      "grad_norm": 0.013317128643393517,
+      "learning_rate": 0.0023403908794788274,
+      "loss": 0.8028,
+      "step": 408
+    },
+    {
+      "epoch": 0.8989010989010989,
+      "grad_norm": 0.008422832936048508,
+      "learning_rate": 0.0023379478827361562,
+      "loss": 0.843,
+      "step": 409
+    },
+    {
+      "epoch": 0.9010989010989011,
+      "grad_norm": 0.010878072120249271,
+      "learning_rate": 0.0023355048859934856,
+      "loss": 0.7196,
+      "step": 410
+    },
+    {
+      "epoch": 0.9032967032967033,
+      "grad_norm": 0.005039801821112633,
+      "learning_rate": 0.0023330618892508145,
+      "loss": 0.8954,
+      "step": 411
+    },
+    {
+      "epoch": 0.9054945054945055,
+      "grad_norm": 0.006794814486056566,
+      "learning_rate": 0.0023306188925081434,
+      "loss": 0.8908,
+      "step": 412
+    },
+    {
+      "epoch": 0.9076923076923077,
+      "grad_norm": 0.005690097343176603,
+      "learning_rate": 0.0023281758957654723,
+      "loss": 0.8156,
+      "step": 413
+    },
+    {
+      "epoch": 0.9098901098901099,
+      "grad_norm": 0.00880100205540657,
+      "learning_rate": 0.0023257328990228016,
+      "loss": 0.8004,
+      "step": 414
+    },
+    {
+      "epoch": 0.9120879120879121,
+      "grad_norm": 0.008073105476796627,
+      "learning_rate": 0.0023232899022801305,
+      "loss": 0.8597,
+      "step": 415
+    },
+    {
+      "epoch": 0.9142857142857143,
+      "grad_norm": 0.010123916901648045,
+      "learning_rate": 0.0023208469055374594,
+      "loss": 0.8457,
+      "step": 416
+    },
+    {
+      "epoch": 0.9164835164835164,
+      "grad_norm": 0.0051084947772324085,
+      "learning_rate": 0.0023184039087947883,
+      "loss": 0.8704,
+      "step": 417
+    },
+    {
+      "epoch": 0.9186813186813186,
+      "grad_norm": 0.007338300812989473,
+      "learning_rate": 0.002315960912052117,
+      "loss": 0.8407,
+      "step": 418
+    },
+    {
+      "epoch": 0.9208791208791208,
+      "grad_norm": 0.005260343197733164,
+      "learning_rate": 0.0023135179153094465,
+      "loss": 0.8202,
+      "step": 419
+    },
+    {
+      "epoch": 0.9230769230769231,
+      "grad_norm": 0.0033912193030118942,
+      "learning_rate": 0.0023110749185667754,
+      "loss": 0.7687,
+      "step": 420
+    },
+    {
+      "epoch": 0.9252747252747253,
+      "grad_norm": 0.006196349393576384,
+      "learning_rate": 0.0023086319218241043,
+      "loss": 0.7446,
+      "step": 421
+    },
+    {
+      "epoch": 0.9274725274725275,
+      "grad_norm": 0.004013248719274998,
+      "learning_rate": 0.002306188925081433,
+      "loss": 0.8039,
+      "step": 422
+    },
+    {
+      "epoch": 0.9296703296703297,
+      "grad_norm": 0.006334323901683092,
+      "learning_rate": 0.0023037459283387625,
+      "loss": 0.8203,
+      "step": 423
+    },
+    {
+      "epoch": 0.9318681318681319,
+      "grad_norm": 0.0057995375245809555,
+      "learning_rate": 0.0023013029315960914,
+      "loss": 0.7177,
+      "step": 424
+    },
+    {
+      "epoch": 0.9340659340659341,
+      "grad_norm": 0.008075746707618237,
+      "learning_rate": 0.0022988599348534203,
+      "loss": 0.7109,
+      "step": 425
+    },
+    {
+      "epoch": 0.9362637362637363,
+      "grad_norm": 0.006034860387444496,
+      "learning_rate": 0.002296416938110749,
+      "loss": 0.8766,
+      "step": 426
+    },
+    {
+      "epoch": 0.9384615384615385,
+      "grad_norm": 0.006750739645212889,
+      "learning_rate": 0.002293973941368078,
+      "loss": 0.9045,
+      "step": 427
+    },
+    {
+      "epoch": 0.9406593406593406,
+      "grad_norm": 0.006651893258094788,
+      "learning_rate": 0.0022915309446254074,
+      "loss": 0.7453,
+      "step": 428
+    },
+    {
+      "epoch": 0.9428571428571428,
+      "grad_norm": 0.005564981140196323,
+      "learning_rate": 0.0022890879478827363,
+      "loss": 0.8544,
+      "step": 429
+    },
+    {
+      "epoch": 0.945054945054945,
+      "grad_norm": 0.004222165327519178,
+      "learning_rate": 0.002286644951140065,
+      "loss": 0.7819,
+      "step": 430
+    },
+    {
+      "epoch": 0.9472527472527472,
+      "grad_norm": 0.0054136947728693485,
+      "learning_rate": 0.002284201954397394,
+      "loss": 0.6238,
+      "step": 431
+    },
+    {
+      "epoch": 0.9494505494505494,
+      "grad_norm": 0.006234462838619947,
+      "learning_rate": 0.0022817589576547234,
+      "loss": 0.7833,
+      "step": 432
+    },
+    {
+      "epoch": 0.9516483516483516,
+      "grad_norm": 0.003745214780792594,
+      "learning_rate": 0.0022793159609120523,
+      "loss": 0.614,
+      "step": 433
+    },
+    {
+      "epoch": 0.9538461538461539,
+      "grad_norm": 0.007407111115753651,
+      "learning_rate": 0.002276872964169381,
+      "loss": 0.9594,
+      "step": 434
+    },
+    {
+      "epoch": 0.9560439560439561,
+      "grad_norm": 0.0052513238042593,
+      "learning_rate": 0.00227442996742671,
+      "loss": 0.6803,
+      "step": 435
+    },
+    {
+      "epoch": 0.9582417582417583,
+      "grad_norm": 0.005091955419629812,
+      "learning_rate": 0.002271986970684039,
+      "loss": 0.7987,
+      "step": 436
+    },
+    {
+      "epoch": 0.9604395604395605,
+      "grad_norm": 0.011887368746101856,
+      "learning_rate": 0.0022695439739413683,
+      "loss": 0.8707,
+      "step": 437
+    },
+    {
+      "epoch": 0.9626373626373627,
+      "grad_norm": 0.010631270706653595,
+      "learning_rate": 0.002267100977198697,
+      "loss": 0.7637,
+      "step": 438
+    },
+    {
+      "epoch": 0.9648351648351648,
+      "grad_norm": 0.007365461904555559,
+      "learning_rate": 0.002264657980456026,
+      "loss": 0.8537,
+      "step": 439
+    },
+    {
+      "epoch": 0.967032967032967,
+      "grad_norm": 0.005710619036108255,
+      "learning_rate": 0.002262214983713355,
+      "loss": 0.7202,
+      "step": 440
+    },
+    {
+      "epoch": 0.9692307692307692,
+      "grad_norm": 0.006766113452613354,
+      "learning_rate": 0.0022597719869706843,
+      "loss": 0.8359,
+      "step": 441
+    },
+    {
+      "epoch": 0.9714285714285714,
+      "grad_norm": 0.007542749866843224,
+      "learning_rate": 0.002257328990228013,
+      "loss": 0.7802,
+      "step": 442
+    },
+    {
+      "epoch": 0.9736263736263736,
+      "grad_norm": 0.027223795652389526,
+      "learning_rate": 0.002254885993485342,
+      "loss": 0.6795,
+      "step": 443
+    },
+    {
+      "epoch": 0.9758241758241758,
+      "grad_norm": 0.006975453346967697,
+      "learning_rate": 0.002252442996742671,
+      "loss": 0.8093,
+      "step": 444
+    },
+    {
+      "epoch": 0.978021978021978,
+      "grad_norm": 0.0050148023292422295,
+      "learning_rate": 0.0022500000000000003,
+      "loss": 0.747,
+      "step": 445
+    },
+    {
+      "epoch": 0.9802197802197802,
+      "grad_norm": 0.00863489881157875,
+      "learning_rate": 0.002247557003257329,
+      "loss": 0.848,
+      "step": 446
+    },
+    {
+      "epoch": 0.9824175824175824,
+      "grad_norm": 0.006560338661074638,
+      "learning_rate": 0.002245114006514658,
+      "loss": 0.662,
+      "step": 447
+    },
+    {
+      "epoch": 0.9846153846153847,
+      "grad_norm": 0.004969564266502857,
+      "learning_rate": 0.002242671009771987,
+      "loss": 0.887,
+      "step": 448
+    },
+    {
+      "epoch": 0.9868131868131869,
+      "grad_norm": 0.006444138940423727,
+      "learning_rate": 0.002240228013029316,
+      "loss": 0.7801,
+      "step": 449
+    },
+    {
+      "epoch": 0.989010989010989,
+      "grad_norm": 0.0053611197508871555,
+      "learning_rate": 0.002237785016286645,
+      "loss": 0.6369,
+      "step": 450
+    },
+    {
+      "epoch": 0.9912087912087912,
+      "grad_norm": 0.0063239336013793945,
+      "learning_rate": 0.002235342019543974,
+      "loss": 0.7867,
+      "step": 451
+    },
+    {
+      "epoch": 0.9934065934065934,
+      "grad_norm": 0.007880776189267635,
+      "learning_rate": 0.002232899022801303,
+      "loss": 0.9698,
+      "step": 452
+    },
+    {
+      "epoch": 0.9956043956043956,
+      "grad_norm": 0.01751020736992359,
+      "learning_rate": 0.002230456026058632,
+      "loss": 0.7916,
+      "step": 453
+    },
+    {
+      "epoch": 0.9978021978021978,
+      "grad_norm": 0.007318670861423016,
+      "learning_rate": 0.002228013029315961,
+      "loss": 0.7306,
+      "step": 454
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.008588163182139397,
+      "learning_rate": 0.00222557003257329,
+      "loss": 0.7054,
+      "step": 455
+    },
+    {
+      "epoch": 1.0021978021978022,
+      "grad_norm": 0.007557283621281385,
+      "learning_rate": 0.002223127035830619,
+      "loss": 0.678,
+      "step": 456
+    },
+    {
+      "epoch": 1.0043956043956044,
+      "grad_norm": 0.009741372428834438,
+      "learning_rate": 0.002220684039087948,
+      "loss": 0.8145,
+      "step": 457
+    },
+    {
+      "epoch": 1.0065934065934066,
+      "grad_norm": 0.00560142332687974,
+      "learning_rate": 0.0022182410423452767,
+      "loss": 0.7984,
+      "step": 458
+    },
+    {
+      "epoch": 1.0087912087912088,
+      "grad_norm": 0.00701606972143054,
+      "learning_rate": 0.002215798045602606,
+      "loss": 0.8294,
+      "step": 459
+    },
+    {
+      "epoch": 1.010989010989011,
+      "grad_norm": 0.03609955310821533,
+      "learning_rate": 0.002213355048859935,
+      "loss": 0.8912,
+      "step": 460
+    },
+    {
+      "epoch": 1.0131868131868131,
+      "grad_norm": 0.00525628961622715,
+      "learning_rate": 0.002210912052117264,
+      "loss": 0.8167,
+      "step": 461
+    },
+    {
+      "epoch": 1.0153846153846153,
+      "grad_norm": 0.008124853484332561,
+      "learning_rate": 0.0022084690553745927,
+      "loss": 0.9082,
+      "step": 462
+    },
+    {
+      "epoch": 1.0175824175824175,
+      "grad_norm": 0.016545111313462257,
+      "learning_rate": 0.002206026058631922,
+      "loss": 0.7483,
+      "step": 463
+    },
+    {
+      "epoch": 1.0197802197802197,
+      "grad_norm": 0.006014868151396513,
+      "learning_rate": 0.002203583061889251,
+      "loss": 0.7543,
+      "step": 464
+    },
+    {
+      "epoch": 1.021978021978022,
+      "grad_norm": 0.007746729999780655,
+      "learning_rate": 0.00220114006514658,
+      "loss": 0.7237,
+      "step": 465
+    },
+    {
+      "epoch": 1.024175824175824,
+      "grad_norm": 0.00702316639944911,
+      "learning_rate": 0.0021986970684039087,
+      "loss": 0.7028,
+      "step": 466
+    },
+    {
+      "epoch": 1.0263736263736263,
+      "grad_norm": 0.004771884065121412,
+      "learning_rate": 0.0021962540716612376,
+      "loss": 0.7504,
+      "step": 467
+    },
+    {
+      "epoch": 1.0285714285714285,
+      "grad_norm": 0.003978740889579058,
+      "learning_rate": 0.002193811074918567,
+      "loss": 0.8414,
+      "step": 468
+    },
+    {
+      "epoch": 1.0307692307692307,
+      "grad_norm": 0.006275990977883339,
+      "learning_rate": 0.002191368078175896,
+      "loss": 0.7502,
+      "step": 469
+    },
+    {
+      "epoch": 1.032967032967033,
+      "grad_norm": 0.003587264334782958,
+      "learning_rate": 0.0021889250814332247,
+      "loss": 0.7778,
+      "step": 470
+    },
+    {
+      "epoch": 1.0351648351648353,
+      "grad_norm": 0.007689167745411396,
+      "learning_rate": 0.0021864820846905536,
+      "loss": 0.753,
+      "step": 471
+    },
+    {
+      "epoch": 1.0373626373626375,
+      "grad_norm": 0.0074620554223656654,
+      "learning_rate": 0.002184039087947883,
+      "loss": 0.9767,
+      "step": 472
+    },
+    {
+      "epoch": 1.0395604395604396,
+      "grad_norm": 0.005569789092987776,
+      "learning_rate": 0.002181596091205212,
+      "loss": 0.7148,
+      "step": 473
+    },
+    {
+      "epoch": 1.0417582417582418,
+      "grad_norm": 0.00618066918104887,
+      "learning_rate": 0.0021791530944625407,
+      "loss": 0.8334,
+      "step": 474
+    },
+    {
+      "epoch": 1.043956043956044,
+      "grad_norm": 0.008715665899217129,
+      "learning_rate": 0.0021767100977198696,
+      "loss": 0.8323,
+      "step": 475
+    },
+    {
+      "epoch": 1.0461538461538462,
+      "grad_norm": 0.004187523387372494,
+      "learning_rate": 0.0021742671009771985,
+      "loss": 0.8292,
+      "step": 476
+    },
+    {
+      "epoch": 1.0483516483516484,
+      "grad_norm": 0.009629838168621063,
+      "learning_rate": 0.002171824104234528,
+      "loss": 0.9143,
+      "step": 477
+    },
+    {
+      "epoch": 1.0505494505494506,
+      "grad_norm": 0.00412255572155118,
+      "learning_rate": 0.0021693811074918567,
+      "loss": 0.683,
+      "step": 478
+    },
+    {
+      "epoch": 1.0527472527472528,
+      "grad_norm": 0.010268045589327812,
+      "learning_rate": 0.0021669381107491856,
+      "loss": 0.8336,
+      "step": 479
+    },
+    {
+      "epoch": 1.054945054945055,
+      "grad_norm": 0.010530585423111916,
+      "learning_rate": 0.0021644951140065145,
+      "loss": 1.0178,
+      "step": 480
+    },
+    {
+      "epoch": 1.0571428571428572,
+      "grad_norm": 0.004061115439981222,
+      "learning_rate": 0.002162052117263844,
+      "loss": 0.7162,
+      "step": 481
+    },
+    {
+      "epoch": 1.0593406593406594,
+      "grad_norm": 0.0043199192732572556,
+      "learning_rate": 0.0021596091205211727,
+      "loss": 0.8186,
+      "step": 482
+    },
+    {
+      "epoch": 1.0615384615384615,
+      "grad_norm": 0.00394221069291234,
+      "learning_rate": 0.0021571661237785016,
+      "loss": 0.6313,
+      "step": 483
+    },
+    {
+      "epoch": 1.0637362637362637,
+      "grad_norm": 0.004699647426605225,
+      "learning_rate": 0.0021547231270358305,
+      "loss": 0.7356,
+      "step": 484
+    },
+    {
+      "epoch": 1.065934065934066,
+      "grad_norm": 0.006669661495834589,
+      "learning_rate": 0.0021522801302931594,
+      "loss": 0.8655,
+      "step": 485
+    },
+    {
+      "epoch": 1.0681318681318681,
+      "grad_norm": 0.007851545698940754,
+      "learning_rate": 0.0021498371335504887,
+      "loss": 0.7271,
+      "step": 486
+    },
+    {
+      "epoch": 1.0703296703296703,
+      "grad_norm": 0.0034689181484282017,
+      "learning_rate": 0.0021473941368078176,
+      "loss": 0.6357,
+      "step": 487
+    },
+    {
+      "epoch": 1.0725274725274725,
+      "grad_norm": 0.0074816858395934105,
+      "learning_rate": 0.0021449511400651465,
+      "loss": 0.6178,
+      "step": 488
+    },
+    {
+      "epoch": 1.0747252747252747,
+      "grad_norm": 0.009442142210900784,
+      "learning_rate": 0.0021425081433224754,
+      "loss": 0.7641,
+      "step": 489
+    },
+    {
+      "epoch": 1.0769230769230769,
+      "grad_norm": 0.004966671112924814,
+      "learning_rate": 0.0021400651465798047,
+      "loss": 0.7241,
+      "step": 490
+    },
+    {
+      "epoch": 1.079120879120879,
+      "grad_norm": 0.0073088547214865685,
+      "learning_rate": 0.0021376221498371336,
+      "loss": 0.8868,
+      "step": 491
+    },
+    {
+      "epoch": 1.0813186813186813,
+      "grad_norm": 0.01682806946337223,
+      "learning_rate": 0.0021351791530944625,
+      "loss": 0.9163,
+      "step": 492
+    },
+    {
+      "epoch": 1.0835164835164834,
+      "grad_norm": 0.005333283916115761,
+      "learning_rate": 0.0021327361563517914,
+      "loss": 0.9786,
+      "step": 493
+    },
+    {
+      "epoch": 1.0857142857142856,
+      "grad_norm": 0.004292118363082409,
+      "learning_rate": 0.0021302931596091203,
+      "loss": 0.7282,
+      "step": 494
+    },
+    {
+      "epoch": 1.0879120879120878,
+      "grad_norm": 0.006371375173330307,
+      "learning_rate": 0.0021278501628664496,
+      "loss": 0.787,
+      "step": 495
+    },
+    {
+      "epoch": 1.09010989010989,
+      "grad_norm": 0.021046549081802368,
+      "learning_rate": 0.0021254071661237785,
+      "loss": 0.8755,
+      "step": 496
+    },
+    {
+      "epoch": 1.0923076923076924,
+      "grad_norm": 0.004348130896687508,
+      "learning_rate": 0.0021229641693811074,
+      "loss": 0.8258,
+      "step": 497
+    },
+    {
+      "epoch": 1.0945054945054946,
+      "grad_norm": 0.007690248545259237,
+      "learning_rate": 0.0021205211726384363,
+      "loss": 0.7628,
+      "step": 498
+    },
+    {
+      "epoch": 1.0967032967032968,
+      "grad_norm": 0.004843440372496843,
+      "learning_rate": 0.0021180781758957656,
+      "loss": 0.8445,
+      "step": 499
+    },
+    {
+      "epoch": 1.098901098901099,
+      "grad_norm": 0.008400721475481987,
+      "learning_rate": 0.0021156351791530945,
+      "loss": 0.8733,
+      "step": 500
+    },
+    {
+      "epoch": 1.1010989010989012,
+      "grad_norm": 0.00782290380448103,
+      "learning_rate": 0.0021131921824104234,
+      "loss": 0.7305,
+      "step": 501
+    },
+    {
+      "epoch": 1.1032967032967034,
+      "grad_norm": 0.005395593121647835,
+      "learning_rate": 0.0021107491856677523,
+      "loss": 0.7957,
+      "step": 502
+    },
+    {
+      "epoch": 1.1054945054945056,
+      "grad_norm": 0.006222647149115801,
+      "learning_rate": 0.002108306188925081,
+      "loss": 0.9277,
+      "step": 503
+    },
+    {
+      "epoch": 1.1076923076923078,
+      "grad_norm": 0.00544757442548871,
+      "learning_rate": 0.0021058631921824105,
+      "loss": 0.8091,
+      "step": 504
+    },
+    {
+      "epoch": 1.10989010989011,
+      "grad_norm": 0.0044324686750769615,
+      "learning_rate": 0.0021034201954397394,
+      "loss": 0.8378,
+      "step": 505
+    },
+    {
+      "epoch": 1.1120879120879121,
+      "grad_norm": 0.00540689192712307,
+      "learning_rate": 0.0021009771986970683,
+      "loss": 0.7995,
+      "step": 506
+    },
+    {
+      "epoch": 1.1142857142857143,
+      "grad_norm": 0.01693885773420334,
+      "learning_rate": 0.002098534201954397,
+      "loss": 0.7746,
+      "step": 507
+    },
+    {
+      "epoch": 1.1164835164835165,
+      "grad_norm": 0.011478684842586517,
+      "learning_rate": 0.0020960912052117265,
+      "loss": 0.8072,
+      "step": 508
+    },
+    {
+      "epoch": 1.1186813186813187,
+      "grad_norm": 0.0067499978467822075,
+      "learning_rate": 0.0020936482084690554,
+      "loss": 0.737,
+      "step": 509
+    },
+    {
+      "epoch": 1.120879120879121,
+      "grad_norm": 0.011071654036641121,
+      "learning_rate": 0.0020912052117263843,
+      "loss": 0.8263,
+      "step": 510
+    },
+    {
+      "epoch": 1.123076923076923,
+      "grad_norm": 0.007255037315189838,
+      "learning_rate": 0.002088762214983713,
+      "loss": 1.0584,
+      "step": 511
+    },
+    {
+      "epoch": 1.1252747252747253,
+      "grad_norm": 0.005240347236394882,
+      "learning_rate": 0.0020863192182410425,
+      "loss": 0.8019,
+      "step": 512
+    },
+    {
+      "epoch": 1.1274725274725275,
+      "grad_norm": 0.005220314487814903,
+      "learning_rate": 0.0020838762214983714,
+      "loss": 0.9153,
+      "step": 513
+    },
+    {
+      "epoch": 1.1296703296703297,
+      "grad_norm": 0.008753658272325993,
+      "learning_rate": 0.0020814332247557003,
+      "loss": 0.7054,
+      "step": 514
+    },
+    {
+      "epoch": 1.1318681318681318,
+      "grad_norm": 0.0051744068041443825,
+      "learning_rate": 0.002078990228013029,
+      "loss": 0.8145,
+      "step": 515
+    },
+    {
+      "epoch": 1.134065934065934,
+      "grad_norm": 0.008236348628997803,
+      "learning_rate": 0.002076547231270358,
+      "loss": 0.7261,
+      "step": 516
+    },
+    {
+      "epoch": 1.1362637362637362,
+      "grad_norm": 0.005556794814765453,
+      "learning_rate": 0.0020741042345276874,
+      "loss": 0.6776,
+      "step": 517
+    },
+    {
+      "epoch": 1.1384615384615384,
+      "grad_norm": 0.009150075726211071,
+      "learning_rate": 0.0020716612377850163,
+      "loss": 0.8246,
+      "step": 518
+    },
+    {
+      "epoch": 1.1406593406593406,
+      "grad_norm": 0.006091612856835127,
+      "learning_rate": 0.002069218241042345,
+      "loss": 0.7277,
+      "step": 519
+    },
+    {
+      "epoch": 1.1428571428571428,
+      "grad_norm": 0.005581388715654612,
+      "learning_rate": 0.002066775244299674,
+      "loss": 0.7374,
+      "step": 520
+    },
+    {
+      "epoch": 1.145054945054945,
+      "grad_norm": 0.00604350958019495,
+      "learning_rate": 0.0020643322475570034,
+      "loss": 0.9372,
+      "step": 521
+    },
+    {
+      "epoch": 1.1472527472527472,
+      "grad_norm": 0.007309609092772007,
+      "learning_rate": 0.0020618892508143323,
+      "loss": 0.9329,
+      "step": 522
+    },
+    {
+      "epoch": 1.1494505494505494,
+      "grad_norm": 0.012178574688732624,
+      "learning_rate": 0.002059446254071661,
+      "loss": 0.6452,
+      "step": 523
+    },
+    {
+      "epoch": 1.1516483516483516,
+      "grad_norm": 0.005498092155903578,
+      "learning_rate": 0.00205700325732899,
+      "loss": 0.8054,
+      "step": 524
+    },
+    {
+      "epoch": 1.1538461538461537,
+      "grad_norm": 0.003116584848612547,
+      "learning_rate": 0.002054560260586319,
+      "loss": 0.7231,
+      "step": 525
+    },
+    {
+      "epoch": 1.156043956043956,
+      "grad_norm": 0.00526055321097374,
+      "learning_rate": 0.0020521172638436483,
+      "loss": 0.677,
+      "step": 526
+    },
+    {
+      "epoch": 1.1582417582417581,
+      "grad_norm": 0.004831845872104168,
+      "learning_rate": 0.002049674267100977,
+      "loss": 0.8055,
+      "step": 527
+    },
+    {
+      "epoch": 1.1604395604395605,
+      "grad_norm": 0.008136963471770287,
+      "learning_rate": 0.002047231270358306,
+      "loss": 0.8765,
+      "step": 528
+    },
+    {
+      "epoch": 1.1626373626373627,
+      "grad_norm": 0.004021990578621626,
+      "learning_rate": 0.002044788273615635,
+      "loss": 0.8811,
+      "step": 529
+    },
+    {
+      "epoch": 1.164835164835165,
+      "grad_norm": 0.004229038488119841,
+      "learning_rate": 0.0020423452768729643,
+      "loss": 0.7763,
+      "step": 530
+    },
+    {
+      "epoch": 1.167032967032967,
+      "grad_norm": 0.007889488711953163,
+      "learning_rate": 0.002039902280130293,
+      "loss": 0.818,
+      "step": 531
+    },
+    {
+      "epoch": 1.1692307692307693,
+      "grad_norm": 0.007561609148979187,
+      "learning_rate": 0.002037459283387622,
+      "loss": 1.068,
+      "step": 532
+    },
+    {
+      "epoch": 1.1714285714285715,
+      "grad_norm": 0.004542628303170204,
+      "learning_rate": 0.002035016286644951,
+      "loss": 0.5292,
+      "step": 533
+    },
+    {
+      "epoch": 1.1736263736263737,
+      "grad_norm": 0.004814864136278629,
+      "learning_rate": 0.00203257328990228,
+      "loss": 1.047,
+      "step": 534
+    },
+    {
+      "epoch": 1.1758241758241759,
+      "grad_norm": 0.006265121046453714,
+      "learning_rate": 0.0020301302931596092,
+      "loss": 0.8699,
+      "step": 535
+    },
+    {
+      "epoch": 1.178021978021978,
+      "grad_norm": 0.00420979131013155,
+      "learning_rate": 0.002027687296416938,
+      "loss": 0.7156,
+      "step": 536
+    },
+    {
+      "epoch": 1.1802197802197802,
+      "grad_norm": 0.005305146332830191,
+      "learning_rate": 0.002025244299674267,
+      "loss": 0.8771,
+      "step": 537
+    },
+    {
+      "epoch": 1.1824175824175824,
+      "grad_norm": 0.007377329282462597,
+      "learning_rate": 0.002022801302931596,
+      "loss": 0.7991,
+      "step": 538
+    },
+    {
+      "epoch": 1.1846153846153846,
+      "grad_norm": 0.006972169503569603,
+      "learning_rate": 0.0020203583061889252,
+      "loss": 1.0316,
+      "step": 539
+    },
+    {
+      "epoch": 1.1868131868131868,
+      "grad_norm": 0.005346004385501146,
+      "learning_rate": 0.002017915309446254,
+      "loss": 0.7967,
+      "step": 540
+    },
+    {
+      "epoch": 1.189010989010989,
+      "grad_norm": 0.006548120174556971,
+      "learning_rate": 0.002015472312703583,
+      "loss": 0.7766,
+      "step": 541
+    },
+    {
+      "epoch": 1.1912087912087912,
+      "grad_norm": 0.008813339285552502,
+      "learning_rate": 0.002013029315960912,
+      "loss": 0.8156,
+      "step": 542
+    },
+    {
+      "epoch": 1.1934065934065934,
+      "grad_norm": 0.004855050705373287,
+      "learning_rate": 0.002010586319218241,
+      "loss": 0.7945,
+      "step": 543
+    },
+    {
+      "epoch": 1.1956043956043956,
+      "grad_norm": 0.006277537439018488,
+      "learning_rate": 0.00200814332247557,
+      "loss": 0.9043,
+      "step": 544
+    },
+    {
+      "epoch": 1.1978021978021978,
+      "grad_norm": 0.008544104173779488,
+      "learning_rate": 0.002005700325732899,
+      "loss": 0.755,
+      "step": 545
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.0056436085142195225,
+      "learning_rate": 0.002003257328990228,
+      "loss": 0.7214,
+      "step": 546
+    },
+    {
+      "epoch": 1.2021978021978021,
+      "grad_norm": 0.009024756960570812,
+      "learning_rate": 0.002000814332247557,
+      "loss": 0.7934,
+      "step": 547
+    },
+    {
+      "epoch": 1.2043956043956043,
+      "grad_norm": 0.005294786300510168,
+      "learning_rate": 0.001998371335504886,
+      "loss": 0.8255,
+      "step": 548
+    },
+    {
+      "epoch": 1.2065934065934065,
+      "grad_norm": 0.008838862180709839,
+      "learning_rate": 0.001995928338762215,
+      "loss": 0.7054,
+      "step": 549
+    },
+    {
+      "epoch": 1.2087912087912087,
+      "grad_norm": 0.006325377617031336,
+      "learning_rate": 0.001993485342019544,
+      "loss": 0.6567,
+      "step": 550
+    },
+    {
+      "epoch": 1.210989010989011,
+      "grad_norm": 0.0052367993630468845,
+      "learning_rate": 0.001991042345276873,
+      "loss": 0.723,
+      "step": 551
+    },
+    {
+      "epoch": 1.213186813186813,
+      "grad_norm": 0.004854496568441391,
+      "learning_rate": 0.0019885993485342017,
+      "loss": 0.8147,
+      "step": 552
+    },
+    {
+      "epoch": 1.2153846153846155,
+      "grad_norm": 0.004286590032279491,
+      "learning_rate": 0.001986156351791531,
+      "loss": 0.7623,
+      "step": 553
+    },
+    {
+      "epoch": 1.2175824175824177,
+      "grad_norm": 0.0070112780667841434,
+      "learning_rate": 0.00198371335504886,
+      "loss": 0.7332,
+      "step": 554
+    },
+    {
+      "epoch": 1.2197802197802199,
+      "grad_norm": 0.009733320213854313,
+      "learning_rate": 0.001981270358306189,
+      "loss": 1.0242,
+      "step": 555
+    },
+    {
+      "epoch": 1.221978021978022,
+      "grad_norm": 0.004819449037313461,
+      "learning_rate": 0.0019788273615635177,
+      "loss": 0.794,
+      "step": 556
+    },
+    {
+      "epoch": 1.2241758241758243,
+      "grad_norm": 0.00535667734220624,
+      "learning_rate": 0.001976384364820847,
+      "loss": 0.8002,
+      "step": 557
+    },
+    {
+      "epoch": 1.2263736263736265,
+      "grad_norm": 0.005017950665205717,
+      "learning_rate": 0.001973941368078176,
+      "loss": 0.6509,
+      "step": 558
+    },
+    {
+      "epoch": 1.2285714285714286,
+      "grad_norm": 0.0074426038190722466,
+      "learning_rate": 0.001971498371335505,
+      "loss": 0.7797,
+      "step": 559
+    },
+    {
+      "epoch": 1.2307692307692308,
+      "grad_norm": 0.006603611167520285,
+      "learning_rate": 0.0019690553745928337,
+      "loss": 0.8124,
+      "step": 560
+    },
+    {
+      "epoch": 1.232967032967033,
+      "grad_norm": 0.005945730488747358,
+      "learning_rate": 0.0019666123778501626,
+      "loss": 0.7243,
+      "step": 561
+    },
+    {
+      "epoch": 1.2351648351648352,
+      "grad_norm": 0.0043726409785449505,
+      "learning_rate": 0.001964169381107492,
+      "loss": 0.7504,
+      "step": 562
+    },
+    {
+      "epoch": 1.2373626373626374,
+      "grad_norm": 0.008456443436443806,
+      "learning_rate": 0.001961726384364821,
+      "loss": 0.7951,
+      "step": 563
+    },
+    {
+      "epoch": 1.2395604395604396,
+      "grad_norm": 0.005294432397931814,
+      "learning_rate": 0.0019592833876221497,
+      "loss": 0.9151,
+      "step": 564
+    },
+    {
+      "epoch": 1.2417582417582418,
+      "grad_norm": 0.0055772303603589535,
+      "learning_rate": 0.0019568403908794786,
+      "loss": 0.8505,
+      "step": 565
+    },
+    {
+      "epoch": 1.243956043956044,
+      "grad_norm": 0.0057614631950855255,
+      "learning_rate": 0.001954397394136808,
+      "loss": 0.7938,
+      "step": 566
+    },
+    {
+      "epoch": 1.2461538461538462,
+      "grad_norm": 0.00633889390155673,
+      "learning_rate": 0.0019519543973941368,
+      "loss": 0.9008,
+      "step": 567
+    },
+    {
+      "epoch": 1.2483516483516484,
+      "grad_norm": 0.004668336361646652,
+      "learning_rate": 0.0019495114006514657,
+      "loss": 0.7227,
+      "step": 568
+    },
+    {
+      "epoch": 1.2505494505494505,
+      "grad_norm": 0.004774572793394327,
+      "learning_rate": 0.0019470684039087948,
+      "loss": 0.7082,
+      "step": 569
+    },
+    {
+      "epoch": 1.2527472527472527,
+      "grad_norm": 0.0058392612263560295,
+      "learning_rate": 0.0019446254071661237,
+      "loss": 0.9572,
+      "step": 570
+    },
+    {
+      "epoch": 1.254945054945055,
+      "grad_norm": 0.009508573450148106,
+      "learning_rate": 0.0019421824104234526,
+      "loss": 0.7668,
+      "step": 571
+    },
+    {
+      "epoch": 1.2571428571428571,
+      "grad_norm": 0.005870030727237463,
+      "learning_rate": 0.0019397394136807817,
+      "loss": 0.7377,
+      "step": 572
+    },
+    {
+      "epoch": 1.2593406593406593,
+      "grad_norm": 0.005874353926628828,
+      "learning_rate": 0.0019372964169381106,
+      "loss": 0.8951,
+      "step": 573
+    },
+    {
+      "epoch": 1.2615384615384615,
+      "grad_norm": 0.007902286015450954,
+      "learning_rate": 0.0019348534201954397,
+      "loss": 0.7892,
+      "step": 574
+    },
+    {
+      "epoch": 1.2637362637362637,
+      "grad_norm": 0.00727734062820673,
+      "learning_rate": 0.0019324104234527686,
+      "loss": 0.8229,
+      "step": 575
+    },
+    {
+      "epoch": 1.2659340659340659,
+      "grad_norm": 0.00608306098729372,
+      "learning_rate": 0.0019299674267100977,
+      "loss": 0.8082,
+      "step": 576
+    },
+    {
+      "epoch": 1.268131868131868,
+      "grad_norm": 0.004648597911000252,
+      "learning_rate": 0.0019275244299674266,
+      "loss": 0.8561,
+      "step": 577
+    },
+    {
+      "epoch": 1.2703296703296703,
+      "grad_norm": 0.006742396391928196,
+      "learning_rate": 0.0019250814332247557,
+      "loss": 0.8801,
+      "step": 578
+    },
+    {
+      "epoch": 1.2725274725274724,
+      "grad_norm": 0.004907965660095215,
+      "learning_rate": 0.0019226384364820846,
+      "loss": 0.5761,
+      "step": 579
+    },
+    {
+      "epoch": 1.2747252747252746,
+      "grad_norm": 0.0068021537736058235,
+      "learning_rate": 0.0019201954397394135,
+      "loss": 0.7556,
+      "step": 580
+    },
+    {
+      "epoch": 1.2769230769230768,
+      "grad_norm": 0.006452186498790979,
+      "learning_rate": 0.0019177524429967426,
+      "loss": 0.6664,
+      "step": 581
+    },
+    {
+      "epoch": 1.279120879120879,
+      "grad_norm": 0.006107355933636427,
+      "learning_rate": 0.0019153094462540715,
+      "loss": 0.8627,
+      "step": 582
+    },
+    {
+      "epoch": 1.2813186813186812,
+      "grad_norm": 0.00870822835713625,
+      "learning_rate": 0.0019128664495114006,
+      "loss": 0.7158,
+      "step": 583
+    },
+    {
+      "epoch": 1.2835164835164834,
+      "grad_norm": 0.008205648511648178,
+      "learning_rate": 0.0019104234527687295,
+      "loss": 0.8954,
+      "step": 584
+    },
+    {
+      "epoch": 1.2857142857142856,
+      "grad_norm": 0.00602708337828517,
+      "learning_rate": 0.0019079804560260586,
+      "loss": 0.8402,
+      "step": 585
+    },
+    {
+      "epoch": 1.2879120879120878,
+      "grad_norm": 0.006946600507944822,
+      "learning_rate": 0.0019055374592833875,
+      "loss": 0.7635,
+      "step": 586
+    },
+    {
+      "epoch": 1.2901098901098902,
+      "grad_norm": 0.0071885851211845875,
+      "learning_rate": 0.0019030944625407166,
+      "loss": 0.8512,
+      "step": 587
+    },
+    {
+      "epoch": 1.2923076923076924,
+      "grad_norm": 0.010412588715553284,
+      "learning_rate": 0.0019006514657980455,
+      "loss": 0.7637,
+      "step": 588
+    },
+    {
+      "epoch": 1.2945054945054946,
+      "grad_norm": 0.010670836083590984,
+      "learning_rate": 0.0018982084690553744,
+      "loss": 0.8239,
+      "step": 589
+    },
+    {
+      "epoch": 1.2967032967032968,
+      "grad_norm": 0.003968672826886177,
+      "learning_rate": 0.0018957654723127035,
+      "loss": 0.7782,
+      "step": 590
+    },
+    {
+      "epoch": 1.298901098901099,
+      "grad_norm": 0.00716335978358984,
+      "learning_rate": 0.0018933224755700324,
+      "loss": 0.9863,
+      "step": 591
+    },
+    {
+      "epoch": 1.3010989010989011,
+      "grad_norm": 0.022200487554073334,
+      "learning_rate": 0.0018908794788273615,
+      "loss": 0.9962,
+      "step": 592
+    },
+    {
+      "epoch": 1.3032967032967033,
+      "grad_norm": 0.006314860191196203,
+      "learning_rate": 0.0018884364820846904,
+      "loss": 0.7542,
+      "step": 593
+    },
+    {
+      "epoch": 1.3054945054945055,
+      "grad_norm": 0.009440748021006584,
+      "learning_rate": 0.0018859934853420195,
+      "loss": 0.8878,
+      "step": 594
+    },
+    {
+      "epoch": 1.3076923076923077,
+      "grad_norm": 0.0060560135170817375,
+      "learning_rate": 0.0018835504885993484,
+      "loss": 0.8175,
+      "step": 595
+    },
+    {
+      "epoch": 1.30989010989011,
+      "grad_norm": 0.0052330042235553265,
+      "learning_rate": 0.0018811074918566775,
+      "loss": 0.6573,
+      "step": 596
+    },
+    {
+      "epoch": 1.312087912087912,
+      "grad_norm": 0.008602769114077091,
+      "learning_rate": 0.0018786644951140064,
+      "loss": 0.8517,
+      "step": 597
+    },
+    {
+      "epoch": 1.3142857142857143,
+      "grad_norm": 0.0043687112629413605,
+      "learning_rate": 0.0018762214983713355,
+      "loss": 0.6634,
+      "step": 598
+    },
+    {
+      "epoch": 1.3164835164835165,
+      "grad_norm": 0.006922916974872351,
+      "learning_rate": 0.0018737785016286646,
+      "loss": 0.8883,
+      "step": 599
+    },
+    {
+      "epoch": 1.3186813186813187,
+      "grad_norm": 0.010779444128274918,
+      "learning_rate": 0.0018713355048859937,
+      "loss": 0.8518,
+      "step": 600
+    },
+    {
+      "epoch": 1.3208791208791208,
+      "grad_norm": 0.005874649155884981,
+      "learning_rate": 0.0018688925081433226,
+      "loss": 0.8038,
+      "step": 601
+    },
+    {
+      "epoch": 1.323076923076923,
+      "grad_norm": 0.006771536078304052,
+      "learning_rate": 0.0018664495114006517,
+      "loss": 0.748,
+      "step": 602
+    },
+    {
+      "epoch": 1.3252747252747252,
+      "grad_norm": 0.005474047269672155,
+      "learning_rate": 0.0018640065146579806,
+      "loss": 0.774,
+      "step": 603
+    },
+    {
+      "epoch": 1.3274725274725274,
+      "grad_norm": 0.007447028532624245,
+      "learning_rate": 0.0018615635179153097,
+      "loss": 0.9689,
+      "step": 604
+    },
+    {
+      "epoch": 1.3296703296703296,
+      "grad_norm": 0.004450726322829723,
+      "learning_rate": 0.0018591205211726386,
+      "loss": 0.8619,
+      "step": 605
+    },
+    {
+      "epoch": 1.3318681318681318,
+      "grad_norm": 0.013868212699890137,
+      "learning_rate": 0.0018566775244299677,
+      "loss": 0.7902,
+      "step": 606
+    },
+    {
+      "epoch": 1.334065934065934,
+      "grad_norm": 0.005837101023644209,
+      "learning_rate": 0.0018542345276872966,
+      "loss": 0.8536,
+      "step": 607
+    },
+    {
+      "epoch": 1.3362637362637364,
+      "grad_norm": 0.007201184518635273,
+      "learning_rate": 0.0018517915309446257,
+      "loss": 0.7513,
+      "step": 608
+    },
+    {
+      "epoch": 1.3384615384615386,
+      "grad_norm": 0.008391979150474072,
+      "learning_rate": 0.0018493485342019546,
+      "loss": 0.8334,
+      "step": 609
+    },
+    {
+      "epoch": 1.3406593406593408,
+      "grad_norm": 0.0055203745141625404,
+      "learning_rate": 0.0018469055374592835,
+      "loss": 0.8503,
+      "step": 610
+    },
+    {
+      "epoch": 1.342857142857143,
+      "grad_norm": 0.009313449263572693,
+      "learning_rate": 0.0018444625407166126,
+      "loss": 0.8166,
+      "step": 611
+    },
+    {
+      "epoch": 1.3450549450549452,
+      "grad_norm": 0.009323901496827602,
+      "learning_rate": 0.0018420195439739415,
+      "loss": 0.7405,
+      "step": 612
+    },
+    {
+      "epoch": 1.3472527472527474,
+      "grad_norm": 0.005672655999660492,
+      "learning_rate": 0.0018395765472312706,
+      "loss": 0.7319,
+      "step": 613
+    },
+    {
+      "epoch": 1.3494505494505495,
+      "grad_norm": 0.01280479971319437,
+      "learning_rate": 0.0018371335504885995,
+      "loss": 0.8872,
+      "step": 614
+    },
+    {
+      "epoch": 1.3516483516483517,
+      "grad_norm": 0.004431379027664661,
+      "learning_rate": 0.0018346905537459286,
+      "loss": 0.6325,
+      "step": 615
+    },
+    {
+      "epoch": 1.353846153846154,
+      "grad_norm": 0.0052208430133759975,
+      "learning_rate": 0.0018322475570032575,
+      "loss": 0.6569,
+      "step": 616
+    },
+    {
+      "epoch": 1.3560439560439561,
+      "grad_norm": 0.004239222034811974,
+      "learning_rate": 0.0018298045602605866,
+      "loss": 0.6626,
+      "step": 617
+    },
+    {
+      "epoch": 1.3582417582417583,
+      "grad_norm": 0.004895912483334541,
+      "learning_rate": 0.0018273615635179155,
+      "loss": 0.8125,
+      "step": 618
+    },
+    {
+      "epoch": 1.3604395604395605,
+      "grad_norm": 0.003910691477358341,
+      "learning_rate": 0.0018249185667752444,
+      "loss": 0.6988,
+      "step": 619
+    },
+    {
+      "epoch": 1.3626373626373627,
+      "grad_norm": 0.006564021576195955,
+      "learning_rate": 0.0018224755700325735,
+      "loss": 0.7752,
+      "step": 620
+    },
+    {
+      "epoch": 1.3648351648351649,
+      "grad_norm": 0.006303540430963039,
+      "learning_rate": 0.0018200325732899024,
+      "loss": 0.8469,
+      "step": 621
+    },
+    {
+      "epoch": 1.367032967032967,
+      "grad_norm": 0.0067222765646874905,
+      "learning_rate": 0.0018175895765472315,
+      "loss": 0.7955,
+      "step": 622
+    },
+    {
+      "epoch": 1.3692307692307693,
+      "grad_norm": 0.0055628628470003605,
+      "learning_rate": 0.0018151465798045604,
+      "loss": 0.8743,
+      "step": 623
+    },
+    {
+      "epoch": 1.3714285714285714,
+      "grad_norm": 0.004700120072811842,
+      "learning_rate": 0.0018127035830618895,
+      "loss": 0.8548,
+      "step": 624
+    },
+    {
+      "epoch": 1.3736263736263736,
+      "grad_norm": 0.004033429082483053,
+      "learning_rate": 0.0018102605863192184,
+      "loss": 0.8134,
+      "step": 625
+    },
+    {
+      "epoch": 1.3758241758241758,
+      "grad_norm": 0.00506405858322978,
+      "learning_rate": 0.0018078175895765475,
+      "loss": 0.6537,
+      "step": 626
+    },
+    {
+      "epoch": 1.378021978021978,
+      "grad_norm": 0.0065112062729895115,
+      "learning_rate": 0.0018053745928338764,
+      "loss": 0.9348,
+      "step": 627
+    },
+    {
+      "epoch": 1.3802197802197802,
+      "grad_norm": 0.006602190900593996,
+      "learning_rate": 0.0018029315960912053,
+      "loss": 0.7267,
+      "step": 628
+    },
+    {
+      "epoch": 1.3824175824175824,
+      "grad_norm": 0.0039736079052090645,
+      "learning_rate": 0.0018004885993485344,
+      "loss": 0.7862,
+      "step": 629
+    },
+    {
+      "epoch": 1.3846153846153846,
+      "grad_norm": 0.004260436166077852,
+      "learning_rate": 0.0017980456026058633,
+      "loss": 0.7624,
+      "step": 630
+    },
+    {
+      "epoch": 1.3868131868131868,
+      "grad_norm": 0.0037388557102531195,
+      "learning_rate": 0.0017956026058631924,
+      "loss": 0.9005,
+      "step": 631
+    },
+    {
+      "epoch": 1.389010989010989,
+      "grad_norm": 0.017267191782593727,
+      "learning_rate": 0.0017931596091205213,
+      "loss": 0.8139,
+      "step": 632
+    },
+    {
+      "epoch": 1.3912087912087912,
+      "grad_norm": 0.00443174596875906,
+      "learning_rate": 0.0017907166123778504,
+      "loss": 0.7062,
+      "step": 633
+    },
+    {
+      "epoch": 1.3934065934065933,
+      "grad_norm": 0.004193700384348631,
+      "learning_rate": 0.0017882736156351793,
+      "loss": 0.8297,
+      "step": 634
+    },
+    {
+      "epoch": 1.3956043956043955,
+      "grad_norm": 0.004873988218605518,
+      "learning_rate": 0.0017858306188925084,
+      "loss": 0.7642,
+      "step": 635
+    },
+    {
+      "epoch": 1.3978021978021977,
+      "grad_norm": 0.006606167182326317,
+      "learning_rate": 0.0017833876221498373,
+      "loss": 0.7487,
+      "step": 636
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.007024228572845459,
+      "learning_rate": 0.0017809446254071664,
+      "loss": 0.7627,
+      "step": 637
+    },
+    {
+      "epoch": 1.402197802197802,
+      "grad_norm": 0.004222473129630089,
+      "learning_rate": 0.0017785016286644953,
+      "loss": 0.8159,
+      "step": 638
+    },
+    {
+      "epoch": 1.4043956043956043,
+      "grad_norm": 0.009550865739583969,
+      "learning_rate": 0.0017760586319218242,
+      "loss": 0.6406,
+      "step": 639
+    },
+    {
+      "epoch": 1.4065934065934065,
+      "grad_norm": 0.004889915231615305,
+      "learning_rate": 0.0017736156351791533,
+      "loss": 0.8161,
+      "step": 640
+    },
+    {
+      "epoch": 1.4087912087912087,
+      "grad_norm": 0.008095957338809967,
+      "learning_rate": 0.0017711726384364822,
+      "loss": 0.8898,
+      "step": 641
+    },
+    {
+      "epoch": 1.4109890109890109,
+      "grad_norm": 0.005690800026059151,
+      "learning_rate": 0.0017687296416938113,
+      "loss": 0.762,
+      "step": 642
+    },
+    {
+      "epoch": 1.413186813186813,
+      "grad_norm": 0.004284800495952368,
+      "learning_rate": 0.0017662866449511402,
+      "loss": 0.6025,
+      "step": 643
+    },
+    {
+      "epoch": 1.4153846153846155,
+      "grad_norm": 0.00535074295476079,
+      "learning_rate": 0.0017638436482084693,
+      "loss": 0.7797,
+      "step": 644
+    },
+    {
+      "epoch": 1.4175824175824177,
+      "grad_norm": 0.007294627372175455,
+      "learning_rate": 0.0017614006514657982,
+      "loss": 1.0026,
+      "step": 645
+    },
+    {
+      "epoch": 1.4197802197802198,
+      "grad_norm": 0.007133009377866983,
+      "learning_rate": 0.0017589576547231273,
+      "loss": 1.0943,
+      "step": 646
+    },
+    {
+      "epoch": 1.421978021978022,
+      "grad_norm": 0.006237710826098919,
+      "learning_rate": 0.0017565146579804562,
+      "loss": 0.7197,
+      "step": 647
+    },
+    {
+      "epoch": 1.4241758241758242,
+      "grad_norm": 0.00745525024831295,
+      "learning_rate": 0.001754071661237785,
+      "loss": 0.9625,
+      "step": 648
+    },
+    {
+      "epoch": 1.4263736263736264,
+      "grad_norm": 0.005896910093724728,
+      "learning_rate": 0.0017516286644951142,
+      "loss": 0.8194,
+      "step": 649
+    },
+    {
+      "epoch": 1.4285714285714286,
+      "grad_norm": 0.005657118279486895,
+      "learning_rate": 0.001749185667752443,
+      "loss": 0.829,
+      "step": 650
+    },
+    {
+      "epoch": 1.4307692307692308,
+      "grad_norm": 0.007129998411983252,
+      "learning_rate": 0.0017467426710097722,
+      "loss": 0.7079,
+      "step": 651
+    },
+    {
+      "epoch": 1.432967032967033,
+      "grad_norm": 0.007198362145572901,
+      "learning_rate": 0.001744299674267101,
+      "loss": 0.7678,
+      "step": 652
+    },
+    {
+      "epoch": 1.4351648351648352,
+      "grad_norm": 0.007235856261104345,
+      "learning_rate": 0.0017418566775244302,
+      "loss": 0.7571,
+      "step": 653
+    },
+    {
+      "epoch": 1.4373626373626374,
+      "grad_norm": 0.0048384820111095905,
+      "learning_rate": 0.001739413680781759,
+      "loss": 0.8571,
+      "step": 654
+    },
+    {
+      "epoch": 1.4395604395604396,
+      "grad_norm": 0.004287781659513712,
+      "learning_rate": 0.0017369706840390882,
+      "loss": 0.6014,
+      "step": 655
+    },
+    {
+      "epoch": 1.4417582417582417,
+      "grad_norm": 0.006446068175137043,
+      "learning_rate": 0.001734527687296417,
+      "loss": 0.8483,
+      "step": 656
+    },
+    {
+      "epoch": 1.443956043956044,
+      "grad_norm": 0.004236277658492327,
+      "learning_rate": 0.001732084690553746,
+      "loss": 0.7459,
+      "step": 657
+    },
+    {
+      "epoch": 1.4461538461538461,
+      "grad_norm": 0.0069379801861941814,
+      "learning_rate": 0.001729641693811075,
+      "loss": 0.8879,
+      "step": 658
+    },
+    {
+      "epoch": 1.4483516483516483,
+      "grad_norm": 0.014682757668197155,
+      "learning_rate": 0.001727198697068404,
+      "loss": 1.0282,
+      "step": 659
+    },
+    {
+      "epoch": 1.4505494505494505,
+      "grad_norm": 0.007400627713650465,
+      "learning_rate": 0.001724755700325733,
+      "loss": 1.0145,
+      "step": 660
+    },
+    {
+      "epoch": 1.4527472527472527,
+      "grad_norm": 0.007536995690315962,
+      "learning_rate": 0.001722312703583062,
+      "loss": 0.833,
+      "step": 661
+    },
+    {
+      "epoch": 1.4549450549450549,
+      "grad_norm": 0.007701661437749863,
+      "learning_rate": 0.001719869706840391,
+      "loss": 0.8752,
+      "step": 662
+    },
+    {
+      "epoch": 1.457142857142857,
+      "grad_norm": 0.006971444468945265,
+      "learning_rate": 0.00171742671009772,
+      "loss": 0.7802,
+      "step": 663
+    },
+    {
+      "epoch": 1.4593406593406593,
+      "grad_norm": 0.004532818216830492,
+      "learning_rate": 0.001714983713355049,
+      "loss": 0.7201,
+      "step": 664
+    },
+    {
+      "epoch": 1.4615384615384617,
+      "grad_norm": 0.005864165723323822,
+      "learning_rate": 0.001712540716612378,
+      "loss": 0.8415,
+      "step": 665
+    },
+    {
+      "epoch": 1.4637362637362639,
+      "grad_norm": 0.009229378774762154,
+      "learning_rate": 0.0017100977198697069,
+      "loss": 0.7036,
+      "step": 666
+    },
+    {
+      "epoch": 1.465934065934066,
+      "grad_norm": 0.006183945573866367,
+      "learning_rate": 0.001707654723127036,
+      "loss": 0.8091,
+      "step": 667
+    },
+    {
+      "epoch": 1.4681318681318682,
+      "grad_norm": 0.0053870719857513905,
+      "learning_rate": 0.0017052117263843649,
+      "loss": 0.6396,
+      "step": 668
+    },
+    {
+      "epoch": 1.4703296703296704,
+      "grad_norm": 0.004009594209492207,
+      "learning_rate": 0.001702768729641694,
+      "loss": 0.7436,
+      "step": 669
+    },
+    {
+      "epoch": 1.4725274725274726,
+      "grad_norm": 0.010637322440743446,
+      "learning_rate": 0.0017003257328990229,
+      "loss": 0.7769,
+      "step": 670
+    },
+    {
+      "epoch": 1.4747252747252748,
+      "grad_norm": 0.005420060828328133,
+      "learning_rate": 0.001697882736156352,
+      "loss": 0.8192,
+      "step": 671
+    },
+    {
+      "epoch": 1.476923076923077,
+      "grad_norm": 0.005665999371558428,
+      "learning_rate": 0.0016954397394136809,
+      "loss": 0.7553,
+      "step": 672
+    },
+    {
+      "epoch": 1.4791208791208792,
+      "grad_norm": 0.005738485138863325,
+      "learning_rate": 0.00169299674267101,
+      "loss": 0.9066,
+      "step": 673
+    },
+    {
+      "epoch": 1.4813186813186814,
+      "grad_norm": 0.005735199898481369,
+      "learning_rate": 0.0016905537459283389,
+      "loss": 0.7222,
+      "step": 674
+    },
+    {
+      "epoch": 1.4835164835164836,
+      "grad_norm": 0.004296691622585058,
+      "learning_rate": 0.001688110749185668,
+      "loss": 0.8444,
+      "step": 675
+    },
+    {
+      "epoch": 1.4857142857142858,
+      "grad_norm": 0.004253997001796961,
+      "learning_rate": 0.0016856677524429969,
+      "loss": 0.7491,
+      "step": 676
+    },
+    {
+      "epoch": 1.487912087912088,
+      "grad_norm": 0.013192079961299896,
+      "learning_rate": 0.0016832247557003258,
+      "loss": 0.7833,
+      "step": 677
+    },
+    {
+      "epoch": 1.4901098901098901,
+      "grad_norm": 0.006565480958670378,
+      "learning_rate": 0.0016807817589576549,
+      "loss": 1.1399,
+      "step": 678
+    },
+    {
+      "epoch": 1.4923076923076923,
+      "grad_norm": 0.006042870692908764,
+      "learning_rate": 0.0016783387622149838,
+      "loss": 0.817,
+      "step": 679
+    },
+    {
+      "epoch": 1.4945054945054945,
+      "grad_norm": 0.006916436366736889,
+      "learning_rate": 0.0016758957654723129,
+      "loss": 0.7051,
+      "step": 680
+    },
+    {
+      "epoch": 1.4967032967032967,
+      "grad_norm": 0.007379190064966679,
+      "learning_rate": 0.0016734527687296418,
+      "loss": 0.6441,
+      "step": 681
+    },
+    {
+      "epoch": 1.498901098901099,
+      "grad_norm": 0.006059740670025349,
+      "learning_rate": 0.0016710097719869709,
+      "loss": 1.008,
+      "step": 682
+    },
+    {
+      "epoch": 1.501098901098901,
+      "grad_norm": 0.00499330647289753,
+      "learning_rate": 0.0016685667752442998,
+      "loss": 0.7763,
+      "step": 683
+    },
+    {
+      "epoch": 1.5032967032967033,
+      "grad_norm": 0.007971618324518204,
+      "learning_rate": 0.0016661237785016289,
+      "loss": 0.8289,
+      "step": 684
+    },
+    {
+      "epoch": 1.5054945054945055,
+      "grad_norm": 0.024213803932070732,
+      "learning_rate": 0.0016636807817589578,
+      "loss": 0.7804,
+      "step": 685
+    },
+    {
+      "epoch": 1.5076923076923077,
+      "grad_norm": 0.007323658559471369,
+      "learning_rate": 0.0016612377850162867,
+      "loss": 0.902,
+      "step": 686
+    },
+    {
+      "epoch": 1.5098901098901099,
+      "grad_norm": 0.006266020238399506,
+      "learning_rate": 0.0016587947882736158,
+      "loss": 0.809,
+      "step": 687
+    },
+    {
+      "epoch": 1.512087912087912,
+      "grad_norm": 0.009785137139260769,
+      "learning_rate": 0.0016563517915309447,
+      "loss": 0.8269,
+      "step": 688
+    },
+    {
+      "epoch": 1.5142857142857142,
+      "grad_norm": 0.009664611890912056,
+      "learning_rate": 0.0016539087947882738,
+      "loss": 0.859,
+      "step": 689
+    },
+    {
+      "epoch": 1.5164835164835164,
+      "grad_norm": 0.00929504819214344,
+      "learning_rate": 0.0016514657980456027,
+      "loss": 0.6739,
+      "step": 690
+    },
+    {
+      "epoch": 1.5186813186813186,
+      "grad_norm": 0.009963120333850384,
+      "learning_rate": 0.0016490228013029318,
+      "loss": 0.6166,
+      "step": 691
+    },
+    {
+      "epoch": 1.5208791208791208,
+      "grad_norm": 0.0072178770788013935,
+      "learning_rate": 0.0016465798045602607,
+      "loss": 0.7303,
+      "step": 692
+    },
+    {
+      "epoch": 1.523076923076923,
+      "grad_norm": 0.006123525556176901,
+      "learning_rate": 0.0016441368078175898,
+      "loss": 0.7966,
+      "step": 693
+    },
+    {
+      "epoch": 1.5252747252747252,
+      "grad_norm": 0.007413902785629034,
+      "learning_rate": 0.0016416938110749187,
+      "loss": 0.8482,
+      "step": 694
+    },
+    {
+      "epoch": 1.5274725274725274,
+      "grad_norm": 0.007545597851276398,
+      "learning_rate": 0.0016392508143322475,
+      "loss": 0.5843,
+      "step": 695
+    },
+    {
+      "epoch": 1.5296703296703296,
+      "grad_norm": 0.020776312798261642,
+      "learning_rate": 0.0016368078175895767,
+      "loss": 0.8032,
+      "step": 696
+    },
+    {
+      "epoch": 1.5318681318681318,
+      "grad_norm": 0.0071355667896568775,
+      "learning_rate": 0.0016343648208469056,
+      "loss": 0.7,
+      "step": 697
+    },
+    {
+      "epoch": 1.534065934065934,
+      "grad_norm": 0.01621301658451557,
+      "learning_rate": 0.0016319218241042347,
+      "loss": 0.7336,
+      "step": 698
+    },
+    {
+      "epoch": 1.5362637362637361,
+      "grad_norm": 0.0064641316421329975,
+      "learning_rate": 0.0016294788273615636,
+      "loss": 0.7182,
+      "step": 699
+    },
+    {
+      "epoch": 1.5384615384615383,
+      "grad_norm": 0.007017648778855801,
+      "learning_rate": 0.0016270358306188927,
+      "loss": 0.813,
+      "step": 700
+    },
+    {
+      "epoch": 1.5406593406593405,
+      "grad_norm": 0.015360669232904911,
+      "learning_rate": 0.0016245928338762216,
+      "loss": 1.0787,
+      "step": 701
+    },
+    {
+      "epoch": 1.5428571428571427,
+      "grad_norm": 0.0046012368984520435,
+      "learning_rate": 0.0016221498371335507,
+      "loss": 0.8408,
+      "step": 702
+    },
+    {
+      "epoch": 1.545054945054945,
+      "grad_norm": 0.007483755704015493,
+      "learning_rate": 0.0016197068403908796,
+      "loss": 0.7676,
+      "step": 703
+    },
+    {
+      "epoch": 1.5472527472527473,
+      "grad_norm": 0.006293860729783773,
+      "learning_rate": 0.0016172638436482087,
+      "loss": 0.7828,
+      "step": 704
+    },
+    {
+      "epoch": 1.5494505494505495,
+      "grad_norm": 0.013528472743928432,
+      "learning_rate": 0.0016148208469055376,
+      "loss": 0.6721,
+      "step": 705
+    },
+    {
+      "epoch": 1.5516483516483517,
+      "grad_norm": 0.008874976076185703,
+      "learning_rate": 0.0016123778501628664,
+      "loss": 1.0364,
+      "step": 706
+    },
+    {
+      "epoch": 1.5538461538461539,
+      "grad_norm": 0.0063871955499053,
+      "learning_rate": 0.0016099348534201956,
+      "loss": 0.6953,
+      "step": 707
+    },
+    {
+      "epoch": 1.556043956043956,
+      "grad_norm": 0.005405406001955271,
+      "learning_rate": 0.0016074918566775244,
+      "loss": 0.824,
+      "step": 708
+    },
+    {
+      "epoch": 1.5582417582417583,
+      "grad_norm": 0.0062369066290557384,
+      "learning_rate": 0.0016050488599348536,
+      "loss": 0.7233,
+      "step": 709
+    },
+    {
+      "epoch": 1.5604395604395604,
+      "grad_norm": 0.005631280597299337,
+      "learning_rate": 0.0016026058631921824,
+      "loss": 0.8888,
+      "step": 710
+    },
+    {
+      "epoch": 1.5626373626373626,
+      "grad_norm": 0.0070943390019237995,
+      "learning_rate": 0.0016001628664495116,
+      "loss": 0.7727,
+      "step": 711
+    },
+    {
+      "epoch": 1.5648351648351648,
+      "grad_norm": 0.0032992265187203884,
+      "learning_rate": 0.0015977198697068404,
+      "loss": 0.6611,
+      "step": 712
+    },
+    {
+      "epoch": 1.567032967032967,
+      "grad_norm": 0.007714875508099794,
+      "learning_rate": 0.0015952768729641696,
+      "loss": 0.7235,
+      "step": 713
+    },
+    {
+      "epoch": 1.5692307692307692,
+      "grad_norm": 0.0056733014062047005,
+      "learning_rate": 0.0015928338762214984,
+      "loss": 0.8259,
+      "step": 714
+    },
+    {
+      "epoch": 1.5714285714285714,
+      "grad_norm": 0.0054899244569242,
+      "learning_rate": 0.0015903908794788273,
+      "loss": 0.8301,
+      "step": 715
+    },
+    {
+      "epoch": 1.5736263736263736,
+      "grad_norm": 0.0062878611497581005,
+      "learning_rate": 0.0015879478827361564,
+      "loss": 0.8716,
+      "step": 716
+    },
+    {
+      "epoch": 1.5758241758241758,
+      "grad_norm": 0.005752011202275753,
+      "learning_rate": 0.0015855048859934853,
+      "loss": 0.7902,
+      "step": 717
+    },
+    {
+      "epoch": 1.578021978021978,
+      "grad_norm": 0.006477563641965389,
+      "learning_rate": 0.0015830618892508145,
+      "loss": 0.814,
+      "step": 718
+    },
+    {
+      "epoch": 1.5802197802197804,
+      "grad_norm": 0.006032464560121298,
+      "learning_rate": 0.0015806188925081433,
+      "loss": 0.7724,
+      "step": 719
+    },
+    {
+      "epoch": 1.5824175824175826,
+      "grad_norm": 0.004271973855793476,
+      "learning_rate": 0.0015781758957654725,
+      "loss": 0.803,
+      "step": 720
+    },
+    {
+      "epoch": 1.5846153846153848,
+      "grad_norm": 0.00732099823653698,
+      "learning_rate": 0.0015757328990228013,
+      "loss": 0.7249,
+      "step": 721
+    },
+    {
+      "epoch": 1.586813186813187,
+      "grad_norm": 0.005539766047149897,
+      "learning_rate": 0.0015732899022801305,
+      "loss": 0.7564,
+      "step": 722
+    },
+    {
+      "epoch": 1.5890109890109891,
+      "grad_norm": 0.0063318246975541115,
+      "learning_rate": 0.0015708469055374593,
+      "loss": 0.8595,
+      "step": 723
+    },
+    {
+      "epoch": 1.5912087912087913,
+      "grad_norm": 0.008011487312614918,
+      "learning_rate": 0.0015684039087947882,
+      "loss": 0.8711,
+      "step": 724
+    },
+    {
+      "epoch": 1.5934065934065935,
+      "grad_norm": 0.007916808128356934,
+      "learning_rate": 0.0015659609120521173,
+      "loss": 0.9448,
+      "step": 725
+    },
+    {
+      "epoch": 1.5956043956043957,
+      "grad_norm": 0.007030043285340071,
+      "learning_rate": 0.0015635179153094462,
+      "loss": 0.9432,
+      "step": 726
+    },
+    {
+      "epoch": 1.597802197802198,
+      "grad_norm": 0.0062830704264342785,
+      "learning_rate": 0.0015610749185667753,
+      "loss": 1.0975,
+      "step": 727
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.005636937916278839,
+      "learning_rate": 0.0015586319218241042,
+      "loss": 0.7979,
+      "step": 728
+    },
+    {
+      "epoch": 1.6021978021978023,
+      "grad_norm": 0.010938640683889389,
+      "learning_rate": 0.0015561889250814333,
+      "loss": 0.7229,
+      "step": 729
+    },
+    {
+      "epoch": 1.6043956043956045,
+      "grad_norm": 0.0077581568621098995,
+      "learning_rate": 0.0015537459283387622,
+      "loss": 0.7452,
+      "step": 730
+    },
+    {
+      "epoch": 1.6065934065934067,
+      "grad_norm": 0.00777699239552021,
+      "learning_rate": 0.0015513029315960913,
+      "loss": 0.758,
+      "step": 731
+    },
+    {
+      "epoch": 1.6087912087912088,
+      "grad_norm": 0.008633402176201344,
+      "learning_rate": 0.0015488599348534202,
+      "loss": 0.8821,
+      "step": 732
+    },
+    {
+      "epoch": 1.610989010989011,
+      "grad_norm": 0.005823650863021612,
+      "learning_rate": 0.0015464169381107491,
+      "loss": 0.7599,
+      "step": 733
+    },
+    {
+      "epoch": 1.6131868131868132,
+      "grad_norm": 0.014622394926846027,
+      "learning_rate": 0.0015439739413680782,
+      "loss": 0.8313,
+      "step": 734
+    },
+    {
+      "epoch": 1.6153846153846154,
+      "grad_norm": 0.008124854415655136,
+      "learning_rate": 0.0015415309446254071,
+      "loss": 0.7565,
+      "step": 735
+    },
+    {
+      "epoch": 1.6175824175824176,
+      "grad_norm": 0.012971216812729836,
+      "learning_rate": 0.0015390879478827362,
+      "loss": 1.1375,
+      "step": 736
+    },
+    {
+      "epoch": 1.6197802197802198,
+      "grad_norm": 0.005248935893177986,
+      "learning_rate": 0.0015366449511400651,
+      "loss": 0.6678,
+      "step": 737
+    },
+    {
+      "epoch": 1.621978021978022,
+      "grad_norm": 0.0070538874715566635,
+      "learning_rate": 0.0015342019543973942,
+      "loss": 0.8565,
+      "step": 738
+    },
+    {
+      "epoch": 1.6241758241758242,
+      "grad_norm": 0.005116897635161877,
+      "learning_rate": 0.0015317589576547231,
+      "loss": 0.8454,
+      "step": 739
+    },
+    {
+      "epoch": 1.6263736263736264,
+      "grad_norm": 0.005649149883538485,
+      "learning_rate": 0.0015293159609120522,
+      "loss": 0.7438,
+      "step": 740
+    },
+    {
+      "epoch": 1.6285714285714286,
+      "grad_norm": 0.0072302985936403275,
+      "learning_rate": 0.0015268729641693811,
+      "loss": 0.7581,
+      "step": 741
+    },
+    {
+      "epoch": 1.6307692307692307,
+      "grad_norm": 0.0055454568937420845,
+      "learning_rate": 0.0015244299674267102,
+      "loss": 0.6516,
+      "step": 742
+    },
+    {
+      "epoch": 1.632967032967033,
+      "grad_norm": 0.004435010254383087,
+      "learning_rate": 0.0015219869706840391,
+      "loss": 0.7468,
+      "step": 743
+    },
+    {
+      "epoch": 1.6351648351648351,
+      "grad_norm": 0.004934519529342651,
+      "learning_rate": 0.001519543973941368,
+      "loss": 0.589,
+      "step": 744
+    },
+    {
+      "epoch": 1.6373626373626373,
+      "grad_norm": 0.008302198722958565,
+      "learning_rate": 0.0015171009771986971,
+      "loss": 0.5109,
+      "step": 745
+    },
+    {
+      "epoch": 1.6395604395604395,
+      "grad_norm": 0.011889921501278877,
+      "learning_rate": 0.001514657980456026,
+      "loss": 0.7002,
+      "step": 746
+    },
+    {
+      "epoch": 1.6417582417582417,
+      "grad_norm": 0.0036850771866738796,
+      "learning_rate": 0.0015122149837133551,
+      "loss": 0.7056,
+      "step": 747
+    },
+    {
+      "epoch": 1.6439560439560439,
+      "grad_norm": 0.007066455204039812,
+      "learning_rate": 0.001509771986970684,
+      "loss": 0.7881,
+      "step": 748
+    },
+    {
+      "epoch": 1.646153846153846,
+      "grad_norm": 0.007527138572186232,
+      "learning_rate": 0.0015073289902280131,
+      "loss": 0.514,
+      "step": 749
+    },
+    {
+      "epoch": 1.6483516483516483,
+      "grad_norm": 0.00601784884929657,
+      "learning_rate": 0.001504885993485342,
+      "loss": 0.6233,
+      "step": 750
+    },
+    {
+      "epoch": 1.6505494505494505,
+      "grad_norm": 0.004565734416246414,
+      "learning_rate": 0.0015024429967426711,
+      "loss": 0.8967,
+      "step": 751
+    },
+    {
+      "epoch": 1.6527472527472526,
+      "grad_norm": 0.005635136738419533,
+      "learning_rate": 0.0015,
+      "loss": 0.7588,
+      "step": 752
+    },
+    {
+      "epoch": 1.6549450549450548,
+      "grad_norm": 0.003977995365858078,
+      "learning_rate": 0.001497557003257329,
+      "loss": 0.7704,
+      "step": 753
+    },
+    {
+      "epoch": 1.657142857142857,
+      "grad_norm": 0.012412186712026596,
+      "learning_rate": 0.001495114006514658,
+      "loss": 0.8303,
+      "step": 754
+    },
+    {
+      "epoch": 1.6593406593406592,
+      "grad_norm": 0.015570256859064102,
+      "learning_rate": 0.001492671009771987,
+      "loss": 0.6769,
+      "step": 755
+    },
+    {
+      "epoch": 1.6615384615384614,
+      "grad_norm": 0.007063649129122496,
+      "learning_rate": 0.001490228013029316,
+      "loss": 0.8403,
+      "step": 756
+    },
+    {
+      "epoch": 1.6637362637362636,
+      "grad_norm": 0.004158693831413984,
+      "learning_rate": 0.001487785016286645,
+      "loss": 0.7962,
+      "step": 757
+    },
+    {
+      "epoch": 1.6659340659340658,
+      "grad_norm": 0.008881604298949242,
+      "learning_rate": 0.001485342019543974,
+      "loss": 0.8666,
+      "step": 758
+    },
+    {
+      "epoch": 1.668131868131868,
+      "grad_norm": 0.010602894239127636,
+      "learning_rate": 0.001482899022801303,
+      "loss": 0.7821,
+      "step": 759
+    },
+    {
+      "epoch": 1.6703296703296702,
+      "grad_norm": 0.021040337160229683,
+      "learning_rate": 0.001480456026058632,
+      "loss": 0.8646,
+      "step": 760
+    },
+    {
+      "epoch": 1.6725274725274726,
+      "grad_norm": 0.00715511804446578,
+      "learning_rate": 0.001478013029315961,
+      "loss": 0.5712,
+      "step": 761
+    },
+    {
+      "epoch": 1.6747252747252748,
+      "grad_norm": 0.006049938499927521,
+      "learning_rate": 0.0014755700325732898,
+      "loss": 0.9412,
+      "step": 762
+    },
+    {
+      "epoch": 1.676923076923077,
+      "grad_norm": 0.009520186111330986,
+      "learning_rate": 0.001473127035830619,
+      "loss": 0.8274,
+      "step": 763
+    },
+    {
+      "epoch": 1.6791208791208792,
+      "grad_norm": 0.005878206808120012,
+      "learning_rate": 0.0014706840390879478,
+      "loss": 0.7771,
+      "step": 764
+    },
+    {
+      "epoch": 1.6813186813186813,
+      "grad_norm": 0.005174014251679182,
+      "learning_rate": 0.001468241042345277,
+      "loss": 0.6417,
+      "step": 765
+    },
+    {
+      "epoch": 1.6835164835164835,
+      "grad_norm": 0.005590626038610935,
+      "learning_rate": 0.0014657980456026058,
+      "loss": 0.721,
+      "step": 766
+    },
+    {
+      "epoch": 1.6857142857142857,
+      "grad_norm": 0.007232989184558392,
+      "learning_rate": 0.001463355048859935,
+      "loss": 0.7787,
+      "step": 767
+    },
+    {
+      "epoch": 1.687912087912088,
+      "grad_norm": 0.017726674675941467,
+      "learning_rate": 0.0014609120521172638,
+      "loss": 0.729,
+      "step": 768
+    },
+    {
+      "epoch": 1.69010989010989,
+      "grad_norm": 0.009804944507777691,
+      "learning_rate": 0.001458469055374593,
+      "loss": 0.9438,
+      "step": 769
+    },
+    {
+      "epoch": 1.6923076923076923,
+      "grad_norm": 0.011409108527004719,
+      "learning_rate": 0.0014560260586319218,
+      "loss": 0.7163,
+      "step": 770
+    },
+    {
+      "epoch": 1.6945054945054945,
+      "grad_norm": 0.003904076525941491,
+      "learning_rate": 0.001453583061889251,
+      "loss": 0.6742,
+      "step": 771
+    },
+    {
+      "epoch": 1.6967032967032967,
+      "grad_norm": 0.00767036247998476,
+      "learning_rate": 0.0014511400651465798,
+      "loss": 0.8549,
+      "step": 772
+    },
+    {
+      "epoch": 1.6989010989010989,
+      "grad_norm": 0.004780925810337067,
+      "learning_rate": 0.0014486970684039087,
+      "loss": 0.8375,
+      "step": 773
+    },
+    {
+      "epoch": 1.701098901098901,
+      "grad_norm": 0.010821767151355743,
+      "learning_rate": 0.0014462540716612378,
+      "loss": 0.8829,
+      "step": 774
+    },
+    {
+      "epoch": 1.7032967032967035,
+      "grad_norm": 0.0054602050222456455,
+      "learning_rate": 0.0014438110749185667,
+      "loss": 0.6869,
+      "step": 775
+    },
+    {
+      "epoch": 1.7054945054945057,
+      "grad_norm": 0.006541840266436338,
+      "learning_rate": 0.0014413680781758958,
+      "loss": 0.7702,
+      "step": 776
+    },
+    {
+      "epoch": 1.7076923076923078,
+      "grad_norm": 0.007716208230704069,
+      "learning_rate": 0.0014389250814332247,
+      "loss": 0.8881,
+      "step": 777
+    },
+    {
+      "epoch": 1.70989010989011,
+      "grad_norm": 0.005454950500279665,
+      "learning_rate": 0.0014364820846905538,
+      "loss": 0.681,
+      "step": 778
+    },
+    {
+      "epoch": 1.7120879120879122,
+      "grad_norm": 0.007018998730927706,
+      "learning_rate": 0.0014340390879478827,
+      "loss": 0.743,
+      "step": 779
+    },
+    {
+      "epoch": 1.7142857142857144,
+      "grad_norm": 0.005593845620751381,
+      "learning_rate": 0.0014315960912052118,
+      "loss": 0.9418,
+      "step": 780
+    },
+    {
+      "epoch": 1.7164835164835166,
+      "grad_norm": 0.007811214309185743,
+      "learning_rate": 0.0014291530944625407,
+      "loss": 0.6699,
+      "step": 781
+    },
+    {
+      "epoch": 1.7186813186813188,
+      "grad_norm": 0.007759306114166975,
+      "learning_rate": 0.0014267100977198696,
+      "loss": 0.804,
+      "step": 782
+    },
+    {
+      "epoch": 1.720879120879121,
+      "grad_norm": 0.011205512098968029,
+      "learning_rate": 0.0014242671009771987,
+      "loss": 0.6812,
+      "step": 783
+    },
+    {
+      "epoch": 1.7230769230769232,
+      "grad_norm": 0.005818730220198631,
+      "learning_rate": 0.0014218241042345276,
+      "loss": 0.7903,
+      "step": 784
+    },
+    {
+      "epoch": 1.7252747252747254,
+      "grad_norm": 0.00830228254199028,
+      "learning_rate": 0.0014193811074918567,
+      "loss": 0.8204,
+      "step": 785
+    },
+    {
+      "epoch": 1.7274725274725276,
+      "grad_norm": 0.0036906637251377106,
+      "learning_rate": 0.0014169381107491856,
+      "loss": 0.7313,
+      "step": 786
+    },
+    {
+      "epoch": 1.7296703296703297,
+      "grad_norm": 0.003935527056455612,
+      "learning_rate": 0.0014144951140065147,
+      "loss": 0.7953,
+      "step": 787
+    },
+    {
+      "epoch": 1.731868131868132,
+      "grad_norm": 0.008108691312372684,
+      "learning_rate": 0.0014120521172638436,
+      "loss": 0.798,
+      "step": 788
+    },
+    {
+      "epoch": 1.7340659340659341,
+      "grad_norm": 0.005105239804834127,
+      "learning_rate": 0.0014096091205211727,
+      "loss": 0.7577,
+      "step": 789
+    },
+    {
+      "epoch": 1.7362637362637363,
+      "grad_norm": 0.009799064137041569,
+      "learning_rate": 0.0014071661237785016,
+      "loss": 0.832,
+      "step": 790
+    },
+    {
+      "epoch": 1.7384615384615385,
+      "grad_norm": 0.004479811992496252,
+      "learning_rate": 0.0014047231270358305,
+      "loss": 0.7108,
+      "step": 791
+    },
+    {
+      "epoch": 1.7406593406593407,
+      "grad_norm": 0.006056270562112331,
+      "learning_rate": 0.0014022801302931596,
+      "loss": 0.8878,
+      "step": 792
+    },
+    {
+      "epoch": 1.7428571428571429,
+      "grad_norm": 0.013903256505727768,
+      "learning_rate": 0.0013998371335504885,
+      "loss": 1.0019,
+      "step": 793
+    },
+    {
+      "epoch": 1.745054945054945,
+      "grad_norm": 0.007488286588340998,
+      "learning_rate": 0.0013973941368078176,
+      "loss": 0.8088,
+      "step": 794
+    },
+    {
+      "epoch": 1.7472527472527473,
+      "grad_norm": 0.013243836350739002,
+      "learning_rate": 0.0013949511400651465,
+      "loss": 0.8625,
+      "step": 795
+    },
+    {
+      "epoch": 1.7494505494505495,
+      "grad_norm": 0.0037774548400193453,
+      "learning_rate": 0.0013925081433224756,
+      "loss": 0.795,
+      "step": 796
+    },
+    {
+      "epoch": 1.7516483516483516,
+      "grad_norm": 0.007580338511615992,
+      "learning_rate": 0.0013900651465798045,
+      "loss": 0.7424,
+      "step": 797
+    },
+    {
+      "epoch": 1.7538461538461538,
+      "grad_norm": 0.007595711853355169,
+      "learning_rate": 0.0013876221498371336,
+      "loss": 0.7035,
+      "step": 798
+    },
+    {
+      "epoch": 1.756043956043956,
+      "grad_norm": 0.005280111450701952,
+      "learning_rate": 0.0013851791530944625,
+      "loss": 0.8923,
+      "step": 799
+    },
+    {
+      "epoch": 1.7582417582417582,
+      "grad_norm": 0.005709439050406218,
+      "learning_rate": 0.0013827361563517914,
+      "loss": 0.7865,
+      "step": 800
+    },
+    {
+      "epoch": 1.7604395604395604,
+      "grad_norm": 0.009806826710700989,
+      "learning_rate": 0.0013802931596091205,
+      "loss": 0.864,
+      "step": 801
+    },
+    {
+      "epoch": 1.7626373626373626,
+      "grad_norm": 0.005695877131074667,
+      "learning_rate": 0.0013778501628664494,
+      "loss": 0.7447,
+      "step": 802
+    },
+    {
+      "epoch": 1.7648351648351648,
+      "grad_norm": 0.0077995555475354195,
+      "learning_rate": 0.0013754071661237785,
+      "loss": 0.7108,
+      "step": 803
+    },
+    {
+      "epoch": 1.767032967032967,
+      "grad_norm": 0.005485584028065205,
+      "learning_rate": 0.0013729641693811074,
+      "loss": 0.6509,
+      "step": 804
+    },
+    {
+      "epoch": 1.7692307692307692,
+      "grad_norm": 0.008434266783297062,
+      "learning_rate": 0.0013705211726384365,
+      "loss": 0.6978,
+      "step": 805
+    },
+    {
+      "epoch": 1.7714285714285714,
+      "grad_norm": 0.005174302961677313,
+      "learning_rate": 0.0013680781758957654,
+      "loss": 0.7819,
+      "step": 806
+    },
+    {
+      "epoch": 1.7736263736263735,
+      "grad_norm": 0.00532882334664464,
+      "learning_rate": 0.0013656351791530945,
+      "loss": 0.863,
+      "step": 807
+    },
+    {
+      "epoch": 1.7758241758241757,
+      "grad_norm": 0.0069867526181042194,
+      "learning_rate": 0.0013631921824104234,
+      "loss": 0.7974,
+      "step": 808
+    },
+    {
+      "epoch": 1.778021978021978,
+      "grad_norm": 0.00672357901930809,
+      "learning_rate": 0.0013607491856677525,
+      "loss": 0.6848,
+      "step": 809
+    },
+    {
+      "epoch": 1.7802197802197801,
+      "grad_norm": 0.007489294745028019,
+      "learning_rate": 0.0013583061889250814,
+      "loss": 0.8015,
+      "step": 810
+    },
+    {
+      "epoch": 1.7824175824175823,
+      "grad_norm": 0.004567979834973812,
+      "learning_rate": 0.0013558631921824103,
+      "loss": 0.826,
+      "step": 811
+    },
+    {
+      "epoch": 1.7846153846153845,
+      "grad_norm": 0.004280353896319866,
+      "learning_rate": 0.0013534201954397394,
+      "loss": 0.6264,
+      "step": 812
+    },
+    {
+      "epoch": 1.7868131868131867,
+      "grad_norm": 0.013979708775877953,
+      "learning_rate": 0.0013509771986970683,
+      "loss": 0.6629,
+      "step": 813
+    },
+    {
+      "epoch": 1.7890109890109889,
+      "grad_norm": 0.00883560348302126,
+      "learning_rate": 0.0013485342019543974,
+      "loss": 0.8141,
+      "step": 814
+    },
+    {
+      "epoch": 1.791208791208791,
+      "grad_norm": 0.006740960292518139,
+      "learning_rate": 0.0013460912052117263,
+      "loss": 0.8026,
+      "step": 815
+    },
+    {
+      "epoch": 1.7934065934065933,
+      "grad_norm": 0.006825266405940056,
+      "learning_rate": 0.0013436482084690554,
+      "loss": 0.8551,
+      "step": 816
+    },
+    {
+      "epoch": 1.7956043956043954,
+      "grad_norm": 0.0069427490234375,
+      "learning_rate": 0.0013412052117263843,
+      "loss": 0.7822,
+      "step": 817
+    },
+    {
+      "epoch": 1.7978021978021979,
+      "grad_norm": 0.003786608576774597,
+      "learning_rate": 0.0013387622149837134,
+      "loss": 0.7027,
+      "step": 818
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.004570479970425367,
+      "learning_rate": 0.0013363192182410423,
+      "loss": 0.8338,
+      "step": 819
+    },
+    {
+      "epoch": 1.8021978021978022,
+      "grad_norm": 0.010009765625,
+      "learning_rate": 0.0013338762214983712,
+      "loss": 0.8693,
+      "step": 820
+    },
+    {
+      "epoch": 1.8043956043956044,
+      "grad_norm": 0.005409985315054655,
+      "learning_rate": 0.0013314332247557003,
+      "loss": 0.8033,
+      "step": 821
+    },
+    {
+      "epoch": 1.8065934065934066,
+      "grad_norm": 0.00500554870814085,
+      "learning_rate": 0.0013289902280130292,
+      "loss": 1.0159,
+      "step": 822
+    },
+    {
+      "epoch": 1.8087912087912088,
+      "grad_norm": 0.005421641748398542,
+      "learning_rate": 0.0013265472312703583,
+      "loss": 0.8794,
+      "step": 823
+    },
+    {
+      "epoch": 1.810989010989011,
+      "grad_norm": 0.005239374004304409,
+      "learning_rate": 0.0013241042345276872,
+      "loss": 0.9115,
+      "step": 824
+    },
+    {
+      "epoch": 1.8131868131868132,
+      "grad_norm": 0.004819195717573166,
+      "learning_rate": 0.0013216612377850163,
+      "loss": 0.8482,
+      "step": 825
+    },
+    {
+      "epoch": 1.8153846153846154,
+      "grad_norm": 0.007601325400173664,
+      "learning_rate": 0.0013192182410423452,
+      "loss": 0.7265,
+      "step": 826
+    },
+    {
+      "epoch": 1.8175824175824176,
+      "grad_norm": 0.005019303411245346,
+      "learning_rate": 0.0013167752442996743,
+      "loss": 0.7021,
+      "step": 827
+    },
+    {
+      "epoch": 1.8197802197802198,
+      "grad_norm": 0.004567787516862154,
+      "learning_rate": 0.0013143322475570032,
+      "loss": 0.624,
+      "step": 828
+    },
+    {
+      "epoch": 1.821978021978022,
+      "grad_norm": 0.0073956991545856,
+      "learning_rate": 0.0013118892508143323,
+      "loss": 0.8089,
+      "step": 829
+    },
+    {
+      "epoch": 1.8241758241758241,
+      "grad_norm": 0.010254235938191414,
+      "learning_rate": 0.0013094462540716614,
+      "loss": 0.817,
+      "step": 830
+    },
+    {
+      "epoch": 1.8263736263736263,
+      "grad_norm": 0.0054016392678022385,
+      "learning_rate": 0.0013070032573289903,
+      "loss": 0.7357,
+      "step": 831
+    },
+    {
+      "epoch": 1.8285714285714287,
+      "grad_norm": 0.006629230920225382,
+      "learning_rate": 0.0013045602605863194,
+      "loss": 0.869,
+      "step": 832
+    },
+    {
+      "epoch": 1.830769230769231,
+      "grad_norm": 0.00668733473867178,
+      "learning_rate": 0.0013021172638436483,
+      "loss": 0.7168,
+      "step": 833
+    },
+    {
+      "epoch": 1.8329670329670331,
+      "grad_norm": 0.011696603149175644,
+      "learning_rate": 0.0012996742671009772,
+      "loss": 0.8892,
+      "step": 834
+    },
+    {
+      "epoch": 1.8351648351648353,
+      "grad_norm": 0.011277318932116032,
+      "learning_rate": 0.0012972312703583063,
+      "loss": 0.665,
+      "step": 835
+    },
+    {
+      "epoch": 1.8373626373626375,
+      "grad_norm": 0.005916838068515062,
+      "learning_rate": 0.0012947882736156352,
+      "loss": 0.7094,
+      "step": 836
+    },
+    {
+      "epoch": 1.8395604395604397,
+      "grad_norm": 0.0054554445669054985,
+      "learning_rate": 0.0012923452768729643,
+      "loss": 0.8861,
+      "step": 837
+    },
+    {
+      "epoch": 1.8417582417582419,
+      "grad_norm": 0.008046870119869709,
+      "learning_rate": 0.0012899022801302932,
+      "loss": 0.747,
+      "step": 838
+    },
+    {
+      "epoch": 1.843956043956044,
+      "grad_norm": 0.005640547722578049,
+      "learning_rate": 0.0012874592833876223,
+      "loss": 0.7735,
+      "step": 839
+    },
+    {
+      "epoch": 1.8461538461538463,
+      "grad_norm": 0.00859291572123766,
+      "learning_rate": 0.0012850162866449512,
+      "loss": 0.9981,
+      "step": 840
+    },
+    {
+      "epoch": 1.8483516483516484,
+      "grad_norm": 0.009248429909348488,
+      "learning_rate": 0.0012825732899022803,
+      "loss": 0.7829,
+      "step": 841
+    },
+    {
+      "epoch": 1.8505494505494506,
+      "grad_norm": 0.017375459894537926,
+      "learning_rate": 0.0012801302931596092,
+      "loss": 0.7795,
+      "step": 842
+    },
+    {
+      "epoch": 1.8527472527472528,
+      "grad_norm": 0.008837989531457424,
+      "learning_rate": 0.0012776872964169383,
+      "loss": 0.9272,
+      "step": 843
+    },
+    {
+      "epoch": 1.854945054945055,
+      "grad_norm": 0.007383580785244703,
+      "learning_rate": 0.0012752442996742672,
+      "loss": 0.7797,
+      "step": 844
+    },
+    {
+      "epoch": 1.8571428571428572,
+      "grad_norm": 0.006501143332570791,
+      "learning_rate": 0.001272801302931596,
+      "loss": 0.7555,
+      "step": 845
+    },
+    {
+      "epoch": 1.8593406593406594,
+      "grad_norm": 0.0043167974799871445,
+      "learning_rate": 0.0012703583061889252,
+      "loss": 0.7049,
+      "step": 846
+    },
+    {
+      "epoch": 1.8615384615384616,
+      "grad_norm": 0.005975545383989811,
+      "learning_rate": 0.001267915309446254,
+      "loss": 0.6826,
+      "step": 847
+    },
+    {
+      "epoch": 1.8637362637362638,
+      "grad_norm": 0.0054549043998122215,
+      "learning_rate": 0.0012654723127035832,
+      "loss": 0.8726,
+      "step": 848
+    },
+    {
+      "epoch": 1.865934065934066,
+      "grad_norm": 0.005251723807305098,
+      "learning_rate": 0.001263029315960912,
+      "loss": 0.8681,
+      "step": 849
+    },
+    {
+      "epoch": 1.8681318681318682,
+      "grad_norm": 0.005087817553430796,
+      "learning_rate": 0.0012605863192182412,
+      "loss": 0.8299,
+      "step": 850
+    },
+    {
+      "epoch": 1.8703296703296703,
+      "grad_norm": 0.006871182005852461,
+      "learning_rate": 0.00125814332247557,
+      "loss": 0.7778,
+      "step": 851
+    },
+    {
+      "epoch": 1.8725274725274725,
+      "grad_norm": 0.0058579519391059875,
+      "learning_rate": 0.0012557003257328992,
+      "loss": 0.7716,
+      "step": 852
+    },
+    {
+      "epoch": 1.8747252747252747,
+      "grad_norm": 0.004637099802494049,
+      "learning_rate": 0.001253257328990228,
+      "loss": 0.8819,
+      "step": 853
+    },
+    {
+      "epoch": 1.876923076923077,
+      "grad_norm": 0.005185327026993036,
+      "learning_rate": 0.001250814332247557,
+      "loss": 0.8665,
+      "step": 854
+    },
+    {
+      "epoch": 1.879120879120879,
+      "grad_norm": 0.008584731258451939,
+      "learning_rate": 0.001248371335504886,
+      "loss": 0.7635,
+      "step": 855
+    },
+    {
+      "epoch": 1.8813186813186813,
+      "grad_norm": 0.0060235499404370785,
+      "learning_rate": 0.001245928338762215,
+      "loss": 0.7669,
+      "step": 856
+    },
+    {
+      "epoch": 1.8835164835164835,
+      "grad_norm": 0.007937912829220295,
+      "learning_rate": 0.001243485342019544,
+      "loss": 0.9218,
+      "step": 857
+    },
+    {
+      "epoch": 1.8857142857142857,
+      "grad_norm": 0.007949517108500004,
+      "learning_rate": 0.001241042345276873,
+      "loss": 0.6004,
+      "step": 858
+    },
+    {
+      "epoch": 1.8879120879120879,
+      "grad_norm": 0.007687769830226898,
+      "learning_rate": 0.001238599348534202,
+      "loss": 0.6751,
+      "step": 859
+    },
+    {
+      "epoch": 1.89010989010989,
+      "grad_norm": 0.010476883500814438,
+      "learning_rate": 0.001236156351791531,
+      "loss": 0.908,
+      "step": 860
+    },
+    {
+      "epoch": 1.8923076923076922,
+      "grad_norm": 0.00560098746791482,
+      "learning_rate": 0.00123371335504886,
+      "loss": 0.9082,
+      "step": 861
+    },
+    {
+      "epoch": 1.8945054945054944,
+      "grad_norm": 0.007539708632975817,
+      "learning_rate": 0.001231270358306189,
+      "loss": 0.8013,
+      "step": 862
+    },
+    {
+      "epoch": 1.8967032967032966,
+      "grad_norm": 0.006455851253122091,
+      "learning_rate": 0.0012288273615635179,
+      "loss": 0.7154,
+      "step": 863
+    },
+    {
+      "epoch": 1.8989010989010988,
+      "grad_norm": 0.004388846922665834,
+      "learning_rate": 0.001226384364820847,
+      "loss": 0.7469,
+      "step": 864
+    },
+    {
+      "epoch": 1.901098901098901,
+      "grad_norm": 0.005492613185197115,
+      "learning_rate": 0.0012239413680781759,
+      "loss": 0.9788,
+      "step": 865
+    },
+    {
+      "epoch": 1.9032967032967032,
+      "grad_norm": 0.004889963660389185,
+      "learning_rate": 0.001221498371335505,
+      "loss": 0.6718,
+      "step": 866
+    },
+    {
+      "epoch": 1.9054945054945054,
+      "grad_norm": 0.010122132487595081,
+      "learning_rate": 0.0012190553745928339,
+      "loss": 0.843,
+      "step": 867
+    },
+    {
+      "epoch": 1.9076923076923076,
+      "grad_norm": 0.005783036816865206,
+      "learning_rate": 0.001216612377850163,
+      "loss": 0.8276,
+      "step": 868
+    },
+    {
+      "epoch": 1.9098901098901098,
+      "grad_norm": 0.006605131085962057,
+      "learning_rate": 0.0012141693811074919,
+      "loss": 0.743,
+      "step": 869
+    },
+    {
+      "epoch": 1.912087912087912,
+      "grad_norm": 0.006933028344064951,
+      "learning_rate": 0.001211726384364821,
+      "loss": 0.7962,
+      "step": 870
+    },
+    {
+      "epoch": 1.9142857142857141,
+      "grad_norm": 0.00810010451823473,
+      "learning_rate": 0.0012092833876221499,
+      "loss": 0.6682,
+      "step": 871
+    },
+    {
+      "epoch": 1.9164835164835163,
+      "grad_norm": 0.006695406045764685,
+      "learning_rate": 0.0012068403908794788,
+      "loss": 0.8725,
+      "step": 872
+    },
+    {
+      "epoch": 1.9186813186813185,
+      "grad_norm": 0.006859564688056707,
+      "learning_rate": 0.0012043973941368079,
+      "loss": 1.028,
+      "step": 873
+    },
+    {
+      "epoch": 1.9208791208791207,
+      "grad_norm": 0.004914683755487204,
+      "learning_rate": 0.0012019543973941368,
+      "loss": 0.7552,
+      "step": 874
+    },
+    {
+      "epoch": 1.9230769230769231,
+      "grad_norm": 0.006237984634935856,
+      "learning_rate": 0.0011995114006514659,
+      "loss": 0.7722,
+      "step": 875
+    },
+    {
+      "epoch": 1.9252747252747253,
+      "grad_norm": 0.0058373152278363705,
+      "learning_rate": 0.0011970684039087948,
+      "loss": 0.7728,
+      "step": 876
+    },
+    {
+      "epoch": 1.9274725274725275,
+      "grad_norm": 0.008421655744314194,
+      "learning_rate": 0.0011946254071661239,
+      "loss": 0.7623,
+      "step": 877
+    },
+    {
+      "epoch": 1.9296703296703297,
+      "grad_norm": 0.005711876787245274,
+      "learning_rate": 0.0011921824104234528,
+      "loss": 0.7237,
+      "step": 878
+    },
+    {
+      "epoch": 1.9318681318681319,
+      "grad_norm": 0.00654617277905345,
+      "learning_rate": 0.001189739413680782,
+      "loss": 0.7245,
+      "step": 879
+    },
+    {
+      "epoch": 1.934065934065934,
+      "grad_norm": 0.008015846833586693,
+      "learning_rate": 0.0011872964169381108,
+      "loss": 0.9436,
+      "step": 880
+    },
+    {
+      "epoch": 1.9362637362637363,
+      "grad_norm": 0.004774228669703007,
+      "learning_rate": 0.00118485342019544,
+      "loss": 0.7468,
+      "step": 881
+    },
+    {
+      "epoch": 1.9384615384615385,
+      "grad_norm": 0.007229213137179613,
+      "learning_rate": 0.0011824104234527688,
+      "loss": 0.7011,
+      "step": 882
+    },
+    {
+      "epoch": 1.9406593406593406,
+      "grad_norm": 0.004378260113298893,
+      "learning_rate": 0.0011799674267100977,
+      "loss": 0.8086,
+      "step": 883
+    },
+    {
+      "epoch": 1.9428571428571428,
+      "grad_norm": 0.008132650516927242,
+      "learning_rate": 0.0011775244299674268,
+      "loss": 0.6066,
+      "step": 884
+    },
+    {
+      "epoch": 1.945054945054945,
+      "grad_norm": 0.006358657963573933,
+      "learning_rate": 0.0011750814332247557,
+      "loss": 0.7164,
+      "step": 885
+    },
+    {
+      "epoch": 1.9472527472527472,
+      "grad_norm": 0.006845328025519848,
+      "learning_rate": 0.0011726384364820848,
+      "loss": 0.779,
+      "step": 886
+    },
+    {
+      "epoch": 1.9494505494505494,
+      "grad_norm": 0.004669147077947855,
+      "learning_rate": 0.0011701954397394137,
+      "loss": 0.7884,
+      "step": 887
+    },
+    {
+      "epoch": 1.9516483516483516,
+      "grad_norm": 0.005971880163997412,
+      "learning_rate": 0.0011677524429967428,
+      "loss": 0.693,
+      "step": 888
+    },
+    {
+      "epoch": 1.953846153846154,
+      "grad_norm": 0.0049233222380280495,
+      "learning_rate": 0.0011653094462540717,
+      "loss": 0.9289,
+      "step": 889
+    },
+    {
+      "epoch": 1.9560439560439562,
+      "grad_norm": 0.007622906938195229,
+      "learning_rate": 0.0011628664495114008,
+      "loss": 0.8026,
+      "step": 890
+    },
+    {
+      "epoch": 1.9582417582417584,
+      "grad_norm": 0.006630508694797754,
+      "learning_rate": 0.0011604234527687297,
+      "loss": 0.8196,
+      "step": 891
+    },
+    {
+      "epoch": 1.9604395604395606,
+      "grad_norm": 0.010128123685717583,
+      "learning_rate": 0.0011579804560260586,
+      "loss": 0.877,
+      "step": 892
+    },
+    {
+      "epoch": 1.9626373626373628,
+      "grad_norm": 0.003931071143597364,
+      "learning_rate": 0.0011555374592833877,
+      "loss": 0.8236,
+      "step": 893
+    },
+    {
+      "epoch": 1.964835164835165,
+      "grad_norm": 0.005391143728047609,
+      "learning_rate": 0.0011530944625407166,
+      "loss": 0.7142,
+      "step": 894
+    },
+    {
+      "epoch": 1.9670329670329672,
+      "grad_norm": 0.01106046698987484,
+      "learning_rate": 0.0011506514657980457,
+      "loss": 0.6014,
+      "step": 895
+    },
+    {
+      "epoch": 1.9692307692307693,
+      "grad_norm": 0.006882380228489637,
+      "learning_rate": 0.0011482084690553746,
+      "loss": 0.843,
+      "step": 896
+    },
+    {
+      "epoch": 1.9714285714285715,
+      "grad_norm": 0.009373163804411888,
+      "learning_rate": 0.0011457654723127037,
+      "loss": 0.8658,
+      "step": 897
+    },
+    {
+      "epoch": 1.9736263736263737,
+      "grad_norm": 0.0068388222716748714,
+      "learning_rate": 0.0011433224755700326,
+      "loss": 1.0897,
+      "step": 898
+    },
+    {
+      "epoch": 1.975824175824176,
+      "grad_norm": 0.0170444268733263,
+      "learning_rate": 0.0011408794788273617,
+      "loss": 0.7407,
+      "step": 899
+    },
+    {
+      "epoch": 1.978021978021978,
+      "grad_norm": 0.0072462065145373344,
+      "learning_rate": 0.0011384364820846906,
+      "loss": 0.9923,
+      "step": 900
+    },
+    {
+      "epoch": 1.9802197802197803,
+      "grad_norm": 0.0058096325956285,
+      "learning_rate": 0.0011359934853420195,
+      "loss": 0.7979,
+      "step": 901
+    },
+    {
+      "epoch": 1.9824175824175825,
+      "grad_norm": 0.0054301368072628975,
+      "learning_rate": 0.0011335504885993486,
+      "loss": 0.763,
+      "step": 902
+    },
+    {
+      "epoch": 1.9846153846153847,
+      "grad_norm": 0.012977972626686096,
+      "learning_rate": 0.0011311074918566775,
+      "loss": 0.785,
+      "step": 903
+    },
+    {
+      "epoch": 1.9868131868131869,
+      "grad_norm": 0.007196017075330019,
+      "learning_rate": 0.0011286644951140066,
+      "loss": 0.8723,
+      "step": 904
+    },
+    {
+      "epoch": 1.989010989010989,
+      "grad_norm": 0.014837213791906834,
+      "learning_rate": 0.0011262214983713355,
+      "loss": 0.7698,
+      "step": 905
+    },
+    {
+      "epoch": 1.9912087912087912,
+      "grad_norm": 0.00596653064712882,
+      "learning_rate": 0.0011237785016286646,
+      "loss": 0.7289,
+      "step": 906
+    },
+    {
+      "epoch": 1.9934065934065934,
+      "grad_norm": 0.013924017548561096,
+      "learning_rate": 0.0011213355048859935,
+      "loss": 0.7928,
+      "step": 907
+    },
+    {
+      "epoch": 1.9956043956043956,
+      "grad_norm": 0.004903504159301519,
+      "learning_rate": 0.0011188925081433226,
+      "loss": 0.8796,
+      "step": 908
+    },
+    {
+      "epoch": 1.9978021978021978,
+      "grad_norm": 0.005753195378929377,
+      "learning_rate": 0.0011164495114006515,
+      "loss": 0.6954,
+      "step": 909
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.005795492324978113,
+      "learning_rate": 0.0011140065146579806,
+      "loss": 0.6853,
+      "step": 910
+    },
+    {
+      "epoch": 2.002197802197802,
+      "grad_norm": 0.005363874137401581,
+      "learning_rate": 0.0011115635179153095,
+      "loss": 0.6527,
+      "step": 911
+    },
+    {
+      "epoch": 2.0043956043956044,
+      "grad_norm": 0.007326677907258272,
+      "learning_rate": 0.0011091205211726384,
+      "loss": 0.7132,
+      "step": 912
+    },
+    {
+      "epoch": 2.0065934065934066,
+      "grad_norm": 0.00692111998796463,
+      "learning_rate": 0.0011066775244299675,
+      "loss": 0.739,
+      "step": 913
+    },
+    {
+      "epoch": 2.0087912087912088,
+      "grad_norm": 0.006022145040333271,
+      "learning_rate": 0.0011042345276872964,
+      "loss": 0.7702,
+      "step": 914
+    },
+    {
+      "epoch": 2.010989010989011,
+      "grad_norm": 0.008833156898617744,
+      "learning_rate": 0.0011017915309446255,
+      "loss": 0.6678,
+      "step": 915
+    },
+    {
+      "epoch": 2.013186813186813,
+      "grad_norm": 0.008896805346012115,
+      "learning_rate": 0.0010993485342019544,
+      "loss": 0.7313,
+      "step": 916
+    },
+    {
+      "epoch": 2.0153846153846153,
+      "grad_norm": 0.005788251757621765,
+      "learning_rate": 0.0010969055374592835,
+      "loss": 0.8614,
+      "step": 917
+    },
+    {
+      "epoch": 2.0175824175824175,
+      "grad_norm": 0.01010427437722683,
+      "learning_rate": 0.0010944625407166124,
+      "loss": 0.7095,
+      "step": 918
+    },
+    {
+      "epoch": 2.0197802197802197,
+      "grad_norm": 0.008091473020613194,
+      "learning_rate": 0.0010920195439739415,
+      "loss": 0.6964,
+      "step": 919
+    },
+    {
+      "epoch": 2.021978021978022,
+      "grad_norm": 0.016026495024561882,
+      "learning_rate": 0.0010895765472312704,
+      "loss": 0.9147,
+      "step": 920
+    },
+    {
+      "epoch": 2.024175824175824,
+      "grad_norm": 0.007413438521325588,
+      "learning_rate": 0.0010871335504885993,
+      "loss": 0.8613,
+      "step": 921
+    },
+    {
+      "epoch": 2.0263736263736263,
+      "grad_norm": 0.009813773445785046,
+      "learning_rate": 0.0010846905537459284,
+      "loss": 0.6237,
+      "step": 922
+    },
+    {
+      "epoch": 2.0285714285714285,
+      "grad_norm": 0.00787780899554491,
+      "learning_rate": 0.0010822475570032573,
+      "loss": 0.8025,
+      "step": 923
+    },
+    {
+      "epoch": 2.0307692307692307,
+      "grad_norm": 0.007290148176252842,
+      "learning_rate": 0.0010798045602605864,
+      "loss": 0.6881,
+      "step": 924
+    },
+    {
+      "epoch": 2.032967032967033,
+      "grad_norm": 0.005524398759007454,
+      "learning_rate": 0.0010773615635179153,
+      "loss": 0.835,
+      "step": 925
+    },
+    {
+      "epoch": 2.035164835164835,
+      "grad_norm": 0.007999306544661522,
+      "learning_rate": 0.0010749185667752444,
+      "loss": 0.7636,
+      "step": 926
+    },
+    {
+      "epoch": 2.0373626373626372,
+      "grad_norm": 0.0061143240891397,
+      "learning_rate": 0.0010724755700325733,
+      "loss": 0.8234,
+      "step": 927
+    },
+    {
+      "epoch": 2.0395604395604394,
+      "grad_norm": 0.009486556984484196,
+      "learning_rate": 0.0010700325732899024,
+      "loss": 0.7612,
+      "step": 928
+    },
+    {
+      "epoch": 2.0417582417582416,
+      "grad_norm": 0.0069499812088906765,
+      "learning_rate": 0.0010675895765472313,
+      "loss": 0.6553,
+      "step": 929
+    },
+    {
+      "epoch": 2.043956043956044,
+      "grad_norm": 0.008488238789141178,
+      "learning_rate": 0.0010651465798045602,
+      "loss": 0.7706,
+      "step": 930
+    },
+    {
+      "epoch": 2.046153846153846,
+      "grad_norm": 0.0059787011705338955,
+      "learning_rate": 0.0010627035830618893,
+      "loss": 0.7775,
+      "step": 931
+    },
+    {
+      "epoch": 2.048351648351648,
+      "grad_norm": 0.005956944078207016,
+      "learning_rate": 0.0010602605863192182,
+      "loss": 0.701,
+      "step": 932
+    },
+    {
+      "epoch": 2.0505494505494504,
+      "grad_norm": 0.010180290788412094,
+      "learning_rate": 0.0010578175895765473,
+      "loss": 1.0177,
+      "step": 933
+    },
+    {
+      "epoch": 2.0527472527472526,
+      "grad_norm": 0.004855899140238762,
+      "learning_rate": 0.0010553745928338762,
+      "loss": 0.6864,
+      "step": 934
+    },
+    {
+      "epoch": 2.0549450549450547,
+      "grad_norm": 0.011847643181681633,
+      "learning_rate": 0.0010529315960912053,
+      "loss": 0.8997,
+      "step": 935
+    },
+    {
+      "epoch": 2.057142857142857,
+      "grad_norm": 0.005903572775423527,
+      "learning_rate": 0.0010504885993485342,
+      "loss": 0.8116,
+      "step": 936
+    },
+    {
+      "epoch": 2.059340659340659,
+      "grad_norm": 0.006605369038879871,
+      "learning_rate": 0.0010480456026058633,
+      "loss": 0.7629,
+      "step": 937
+    },
+    {
+      "epoch": 2.0615384615384613,
+      "grad_norm": 0.006202308926731348,
+      "learning_rate": 0.0010456026058631922,
+      "loss": 0.9003,
+      "step": 938
+    },
+    {
+      "epoch": 2.0637362637362635,
+      "grad_norm": 0.008275089785456657,
+      "learning_rate": 0.0010431596091205213,
+      "loss": 0.8815,
+      "step": 939
+    },
+    {
+      "epoch": 2.065934065934066,
+      "grad_norm": 0.004836083855479956,
+      "learning_rate": 0.0010407166123778502,
+      "loss": 0.6472,
+      "step": 940
+    },
+    {
+      "epoch": 2.0681318681318683,
+      "grad_norm": 0.006066104397177696,
+      "learning_rate": 0.001038273615635179,
+      "loss": 0.6389,
+      "step": 941
+    },
+    {
+      "epoch": 2.0703296703296705,
+      "grad_norm": 0.008579510264098644,
+      "learning_rate": 0.0010358306188925082,
+      "loss": 0.8246,
+      "step": 942
+    },
+    {
+      "epoch": 2.0725274725274727,
+      "grad_norm": 0.005505021195858717,
+      "learning_rate": 0.001033387622149837,
+      "loss": 0.8852,
+      "step": 943
+    },
+    {
+      "epoch": 2.074725274725275,
+      "grad_norm": 0.00910259410738945,
+      "learning_rate": 0.0010309446254071662,
+      "loss": 0.6324,
+      "step": 944
+    },
+    {
+      "epoch": 2.076923076923077,
+      "grad_norm": 0.007872145622968674,
+      "learning_rate": 0.001028501628664495,
+      "loss": 0.9053,
+      "step": 945
+    },
+    {
+      "epoch": 2.0791208791208793,
+      "grad_norm": 0.010053028352558613,
+      "learning_rate": 0.0010260586319218242,
+      "loss": 0.8011,
+      "step": 946
+    },
+    {
+      "epoch": 2.0813186813186815,
+      "grad_norm": 0.0072616105899214745,
+      "learning_rate": 0.001023615635179153,
+      "loss": 0.7867,
+      "step": 947
+    },
+    {
+      "epoch": 2.0835164835164837,
+      "grad_norm": 0.011883702129125595,
+      "learning_rate": 0.0010211726384364822,
+      "loss": 0.7382,
+      "step": 948
+    },
+    {
+      "epoch": 2.085714285714286,
+      "grad_norm": 0.008427871391177177,
+      "learning_rate": 0.001018729641693811,
+      "loss": 0.8173,
+      "step": 949
+    },
+    {
+      "epoch": 2.087912087912088,
+      "grad_norm": 0.006745744496583939,
+      "learning_rate": 0.00101628664495114,
+      "loss": 0.6282,
+      "step": 950
+    },
+    {
+      "epoch": 2.0901098901098902,
+      "grad_norm": 0.0054099299013614655,
+      "learning_rate": 0.001013843648208469,
+      "loss": 0.7744,
+      "step": 951
+    },
+    {
+      "epoch": 2.0923076923076924,
+      "grad_norm": 0.012327845208346844,
+      "learning_rate": 0.001011400651465798,
+      "loss": 0.7204,
+      "step": 952
+    },
+    {
+      "epoch": 2.0945054945054946,
+      "grad_norm": 0.004452109336853027,
+      "learning_rate": 0.001008957654723127,
+      "loss": 0.7961,
+      "step": 953
+    },
+    {
+      "epoch": 2.096703296703297,
+      "grad_norm": 0.011333029717206955,
+      "learning_rate": 0.001006514657980456,
+      "loss": 0.7864,
+      "step": 954
+    },
+    {
+      "epoch": 2.098901098901099,
+      "grad_norm": 0.007286337204277515,
+      "learning_rate": 0.001004071661237785,
+      "loss": 0.7571,
+      "step": 955
+    },
+    {
+      "epoch": 2.101098901098901,
+      "grad_norm": 0.007896456867456436,
+      "learning_rate": 0.001001628664495114,
+      "loss": 0.828,
+      "step": 956
+    },
+    {
+      "epoch": 2.1032967032967034,
+      "grad_norm": 0.013499187305569649,
+      "learning_rate": 0.000999185667752443,
+      "loss": 0.8188,
+      "step": 957
+    },
+    {
+      "epoch": 2.1054945054945056,
+      "grad_norm": 0.006730223540216684,
+      "learning_rate": 0.000996742671009772,
+      "loss": 0.751,
+      "step": 958
+    },
+    {
+      "epoch": 2.1076923076923078,
+      "grad_norm": 0.005976951215416193,
+      "learning_rate": 0.0009942996742671008,
+      "loss": 0.6891,
+      "step": 959
+    },
+    {
+      "epoch": 2.10989010989011,
+      "grad_norm": 0.0041262321174144745,
+      "learning_rate": 0.00099185667752443,
+      "loss": 0.8399,
+      "step": 960
+    },
+    {
+      "epoch": 2.112087912087912,
+      "grad_norm": 0.006108524743467569,
+      "learning_rate": 0.0009894136807817588,
+      "loss": 0.7602,
+      "step": 961
+    },
+    {
+      "epoch": 2.1142857142857143,
+      "grad_norm": 0.0057633668184280396,
+      "learning_rate": 0.000986970684039088,
+      "loss": 0.8238,
+      "step": 962
+    },
+    {
+      "epoch": 2.1164835164835165,
+      "grad_norm": 0.005280874669551849,
+      "learning_rate": 0.0009845276872964168,
+      "loss": 0.8521,
+      "step": 963
+    },
+    {
+      "epoch": 2.1186813186813187,
+      "grad_norm": 0.016751151531934738,
+      "learning_rate": 0.000982084690553746,
+      "loss": 0.7729,
+      "step": 964
+    },
+    {
+      "epoch": 2.120879120879121,
+      "grad_norm": 0.007967294193804264,
+      "learning_rate": 0.0009796416938110748,
+      "loss": 0.6138,
+      "step": 965
+    },
+    {
+      "epoch": 2.123076923076923,
+      "grad_norm": 0.012345108203589916,
+      "learning_rate": 0.000977198697068404,
+      "loss": 0.7611,
+      "step": 966
+    },
+    {
+      "epoch": 2.1252747252747253,
+      "grad_norm": 0.00720963953062892,
+      "learning_rate": 0.0009747557003257328,
+      "loss": 0.8995,
+      "step": 967
+    },
+    {
+      "epoch": 2.1274725274725275,
+      "grad_norm": 0.005512729752808809,
+      "learning_rate": 0.0009723127035830618,
+      "loss": 0.5854,
+      "step": 968
+    },
+    {
+      "epoch": 2.1296703296703297,
+      "grad_norm": 0.007667060475796461,
+      "learning_rate": 0.0009698697068403908,
+      "loss": 0.788,
+      "step": 969
+    },
+    {
+      "epoch": 2.131868131868132,
+      "grad_norm": 0.006315926089882851,
+      "learning_rate": 0.0009674267100977198,
+      "loss": 0.7757,
+      "step": 970
+    },
+    {
+      "epoch": 2.134065934065934,
+      "grad_norm": 0.005400488153100014,
+      "learning_rate": 0.0009649837133550488,
+      "loss": 0.7405,
+      "step": 971
+    },
+    {
+      "epoch": 2.1362637362637362,
+      "grad_norm": 0.005479366518557072,
+      "learning_rate": 0.0009625407166123778,
+      "loss": 0.8926,
+      "step": 972
+    },
+    {
+      "epoch": 2.1384615384615384,
+      "grad_norm": 0.005679652560502291,
+      "learning_rate": 0.0009600977198697067,
+      "loss": 0.697,
+      "step": 973
+    },
+    {
+      "epoch": 2.1406593406593406,
+      "grad_norm": 0.007769608870148659,
+      "learning_rate": 0.0009576547231270357,
+      "loss": 0.6211,
+      "step": 974
+    },
+    {
+      "epoch": 2.142857142857143,
+      "grad_norm": 0.008518649265170097,
+      "learning_rate": 0.0009552117263843647,
+      "loss": 0.9518,
+      "step": 975
+    },
+    {
+      "epoch": 2.145054945054945,
+      "grad_norm": 0.005214856471866369,
+      "learning_rate": 0.0009527687296416937,
+      "loss": 0.9265,
+      "step": 976
+    },
+    {
+      "epoch": 2.147252747252747,
+      "grad_norm": 0.006613228935748339,
+      "learning_rate": 0.0009503257328990227,
+      "loss": 0.7941,
+      "step": 977
+    },
+    {
+      "epoch": 2.1494505494505494,
+      "grad_norm": 0.006451161578297615,
+      "learning_rate": 0.0009478827361563517,
+      "loss": 0.8293,
+      "step": 978
+    },
+    {
+      "epoch": 2.1516483516483516,
+      "grad_norm": 0.005191734526306391,
+      "learning_rate": 0.0009454397394136807,
+      "loss": 0.759,
+      "step": 979
+    },
+    {
+      "epoch": 2.1538461538461537,
+      "grad_norm": 0.009721934795379639,
+      "learning_rate": 0.0009429967426710097,
+      "loss": 0.6906,
+      "step": 980
+    },
+    {
+      "epoch": 2.156043956043956,
+      "grad_norm": 0.006205370649695396,
+      "learning_rate": 0.0009405537459283387,
+      "loss": 0.7054,
+      "step": 981
+    },
+    {
+      "epoch": 2.158241758241758,
+      "grad_norm": 0.005210328381508589,
+      "learning_rate": 0.0009381107491856677,
+      "loss": 0.801,
+      "step": 982
+    },
+    {
+      "epoch": 2.1604395604395603,
+      "grad_norm": 0.0052205659449100494,
+      "learning_rate": 0.0009356677524429969,
+      "loss": 0.6577,
+      "step": 983
+    },
+    {
+      "epoch": 2.1626373626373625,
+      "grad_norm": 0.0041876561008393764,
+      "learning_rate": 0.0009332247557003259,
+      "loss": 0.7456,
+      "step": 984
+    },
+    {
+      "epoch": 2.1648351648351647,
+      "grad_norm": 0.005803836043924093,
+      "learning_rate": 0.0009307817589576549,
+      "loss": 0.7639,
+      "step": 985
+    },
+    {
+      "epoch": 2.167032967032967,
+      "grad_norm": 0.006574958562850952,
+      "learning_rate": 0.0009283387622149839,
+      "loss": 0.6896,
+      "step": 986
+    },
+    {
+      "epoch": 2.169230769230769,
+      "grad_norm": 0.005346279591321945,
+      "learning_rate": 0.0009258957654723129,
+      "loss": 0.7361,
+      "step": 987
+    },
+    {
+      "epoch": 2.1714285714285713,
+      "grad_norm": 0.0054899766109883785,
+      "learning_rate": 0.0009234527687296417,
+      "loss": 0.7852,
+      "step": 988
+    },
+    {
+      "epoch": 2.1736263736263735,
+      "grad_norm": 0.005246486980468035,
+      "learning_rate": 0.0009210097719869707,
+      "loss": 0.9217,
+      "step": 989
+    },
+    {
+      "epoch": 2.1758241758241756,
+      "grad_norm": 0.006285570561885834,
+      "learning_rate": 0.0009185667752442997,
+      "loss": 0.8228,
+      "step": 990
+    },
+    {
+      "epoch": 2.178021978021978,
+      "grad_norm": 0.0052154106087982655,
+      "learning_rate": 0.0009161237785016287,
+      "loss": 0.7402,
+      "step": 991
+    },
+    {
+      "epoch": 2.18021978021978,
+      "grad_norm": 0.0061172861605882645,
+      "learning_rate": 0.0009136807817589577,
+      "loss": 0.8305,
+      "step": 992
+    },
+    {
+      "epoch": 2.182417582417582,
+      "grad_norm": 0.005535179283469915,
+      "learning_rate": 0.0009112377850162867,
+      "loss": 0.9058,
+      "step": 993
+    },
+    {
+      "epoch": 2.184615384615385,
+      "grad_norm": 0.007325231563299894,
+      "learning_rate": 0.0009087947882736157,
+      "loss": 0.7526,
+      "step": 994
+    },
+    {
+      "epoch": 2.186813186813187,
+      "grad_norm": 0.004227994941174984,
+      "learning_rate": 0.0009063517915309447,
+      "loss": 0.6607,
+      "step": 995
+    },
+    {
+      "epoch": 2.1890109890109892,
+      "grad_norm": 0.007255845237523317,
+      "learning_rate": 0.0009039087947882737,
+      "loss": 0.766,
+      "step": 996
+    },
+    {
+      "epoch": 2.1912087912087914,
+      "grad_norm": 0.007343007251620293,
+      "learning_rate": 0.0009014657980456026,
+      "loss": 1.002,
+      "step": 997
+    },
+    {
+      "epoch": 2.1934065934065936,
+      "grad_norm": 0.011658702045679092,
+      "learning_rate": 0.0008990228013029316,
+      "loss": 0.75,
+      "step": 998
+    },
+    {
+      "epoch": 2.195604395604396,
+      "grad_norm": 0.004733691923320293,
+      "learning_rate": 0.0008965798045602606,
+      "loss": 0.8513,
+      "step": 999
+    },
+    {
+      "epoch": 2.197802197802198,
+      "grad_norm": 0.009025716222822666,
+      "learning_rate": 0.0008941368078175896,
+      "loss": 0.9192,
+      "step": 1000
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.005283764563500881,
+      "learning_rate": 0.0008916938110749186,
+      "loss": 0.7966,
+      "step": 1001
+    },
+    {
+      "epoch": 2.2021978021978024,
+      "grad_norm": 0.01931913010776043,
+      "learning_rate": 0.0008892508143322476,
+      "loss": 0.6749,
+      "step": 1002
+    },
+    {
+      "epoch": 2.2043956043956046,
+      "grad_norm": 0.009752519428730011,
+      "learning_rate": 0.0008868078175895766,
+      "loss": 0.821,
+      "step": 1003
+    },
+    {
+      "epoch": 2.2065934065934067,
+      "grad_norm": 0.004780410323292017,
+      "learning_rate": 0.0008843648208469056,
+      "loss": 0.5802,
+      "step": 1004
+    },
+    {
+      "epoch": 2.208791208791209,
+      "grad_norm": 0.009423421695828438,
+      "learning_rate": 0.0008819218241042346,
+      "loss": 0.9954,
+      "step": 1005
+    },
+    {
+      "epoch": 2.210989010989011,
+      "grad_norm": 0.0061792549677193165,
+      "learning_rate": 0.0008794788273615636,
+      "loss": 0.8299,
+      "step": 1006
+    },
+    {
+      "epoch": 2.2131868131868133,
+      "grad_norm": 0.01009944174438715,
+      "learning_rate": 0.0008770358306188925,
+      "loss": 0.8153,
+      "step": 1007
+    },
+    {
+      "epoch": 2.2153846153846155,
+      "grad_norm": 0.014022254385054111,
+      "learning_rate": 0.0008745928338762215,
+      "loss": 0.8489,
+      "step": 1008
+    },
+    {
+      "epoch": 2.2175824175824177,
+      "grad_norm": 0.006755861919373274,
+      "learning_rate": 0.0008721498371335505,
+      "loss": 0.8961,
+      "step": 1009
+    },
+    {
+      "epoch": 2.21978021978022,
+      "grad_norm": 0.00796017050743103,
+      "learning_rate": 0.0008697068403908795,
+      "loss": 0.7471,
+      "step": 1010
+    },
+    {
+      "epoch": 2.221978021978022,
+      "grad_norm": 0.004567121155560017,
+      "learning_rate": 0.0008672638436482085,
+      "loss": 0.6743,
+      "step": 1011
+    },
+    {
+      "epoch": 2.2241758241758243,
+      "grad_norm": 0.01266453880816698,
+      "learning_rate": 0.0008648208469055375,
+      "loss": 0.8951,
+      "step": 1012
+    },
+    {
+      "epoch": 2.2263736263736265,
+      "grad_norm": 0.0062808264046907425,
+      "learning_rate": 0.0008623778501628665,
+      "loss": 0.8767,
+      "step": 1013
+    },
+    {
+      "epoch": 2.2285714285714286,
+      "grad_norm": 0.0068176486529409885,
+      "learning_rate": 0.0008599348534201955,
+      "loss": 1.0807,
+      "step": 1014
+    },
+    {
+      "epoch": 2.230769230769231,
+      "grad_norm": 0.008865762501955032,
+      "learning_rate": 0.0008574918566775245,
+      "loss": 0.9914,
+      "step": 1015
+    },
+    {
+      "epoch": 2.232967032967033,
+      "grad_norm": 0.004884714260697365,
+      "learning_rate": 0.0008550488599348534,
+      "loss": 0.8031,
+      "step": 1016
+    },
+    {
+      "epoch": 2.235164835164835,
+      "grad_norm": 0.0053459047339856625,
+      "learning_rate": 0.0008526058631921824,
+      "loss": 0.6822,
+      "step": 1017
+    },
+    {
+      "epoch": 2.2373626373626374,
+      "grad_norm": 0.012971541844308376,
+      "learning_rate": 0.0008501628664495114,
+      "loss": 0.6783,
+      "step": 1018
+    },
+    {
+      "epoch": 2.2395604395604396,
+      "grad_norm": 0.009878777898848057,
+      "learning_rate": 0.0008477198697068404,
+      "loss": 0.7309,
+      "step": 1019
+    },
+    {
+      "epoch": 2.241758241758242,
+      "grad_norm": 0.005557881202548742,
+      "learning_rate": 0.0008452768729641694,
+      "loss": 0.6517,
+      "step": 1020
+    },
+    {
+      "epoch": 2.243956043956044,
+      "grad_norm": 0.01826348528265953,
+      "learning_rate": 0.0008428338762214984,
+      "loss": 0.8205,
+      "step": 1021
+    },
+    {
+      "epoch": 2.246153846153846,
+      "grad_norm": 0.005132089368999004,
+      "learning_rate": 0.0008403908794788274,
+      "loss": 0.7486,
+      "step": 1022
+    },
+    {
+      "epoch": 2.2483516483516484,
+      "grad_norm": 0.007566589396446943,
+      "learning_rate": 0.0008379478827361564,
+      "loss": 0.6971,
+      "step": 1023
+    },
+    {
+      "epoch": 2.2505494505494505,
+      "grad_norm": 0.0070630828849971294,
+      "learning_rate": 0.0008355048859934854,
+      "loss": 0.8193,
+      "step": 1024
+    },
+    {
+      "epoch": 2.2527472527472527,
+      "grad_norm": 0.005031317938119173,
+      "learning_rate": 0.0008330618892508144,
+      "loss": 0.653,
+      "step": 1025
+    },
+    {
+      "epoch": 2.254945054945055,
+      "grad_norm": 0.008693298324942589,
+      "learning_rate": 0.0008306188925081433,
+      "loss": 0.7593,
+      "step": 1026
+    },
+    {
+      "epoch": 2.257142857142857,
+      "grad_norm": 0.005845671519637108,
+      "learning_rate": 0.0008281758957654723,
+      "loss": 0.768,
+      "step": 1027
+    },
+    {
+      "epoch": 2.2593406593406593,
+      "grad_norm": 0.00756713654845953,
+      "learning_rate": 0.0008257328990228013,
+      "loss": 0.81,
+      "step": 1028
+    },
+    {
+      "epoch": 2.2615384615384615,
+      "grad_norm": 0.007276487536728382,
+      "learning_rate": 0.0008232899022801303,
+      "loss": 0.6556,
+      "step": 1029
+    },
+    {
+      "epoch": 2.2637362637362637,
+      "grad_norm": 0.007349386345595121,
+      "learning_rate": 0.0008208469055374593,
+      "loss": 0.8898,
+      "step": 1030
+    },
+    {
+      "epoch": 2.265934065934066,
+      "grad_norm": 0.0061518000438809395,
+      "learning_rate": 0.0008184039087947883,
+      "loss": 0.6442,
+      "step": 1031
+    },
+    {
+      "epoch": 2.268131868131868,
+      "grad_norm": 0.006380500737577677,
+      "learning_rate": 0.0008159609120521173,
+      "loss": 0.8573,
+      "step": 1032
+    },
+    {
+      "epoch": 2.2703296703296703,
+      "grad_norm": 0.005677740555256605,
+      "learning_rate": 0.0008135179153094463,
+      "loss": 0.8188,
+      "step": 1033
+    },
+    {
+      "epoch": 2.2725274725274724,
+      "grad_norm": 0.00739656575024128,
+      "learning_rate": 0.0008110749185667753,
+      "loss": 0.8435,
+      "step": 1034
+    },
+    {
+      "epoch": 2.2747252747252746,
+      "grad_norm": 0.010278203524649143,
+      "learning_rate": 0.0008086319218241043,
+      "loss": 1.0071,
+      "step": 1035
+    },
+    {
+      "epoch": 2.276923076923077,
+      "grad_norm": 0.004747175145894289,
+      "learning_rate": 0.0008061889250814332,
+      "loss": 0.921,
+      "step": 1036
+    },
+    {
+      "epoch": 2.279120879120879,
+      "grad_norm": 0.007145237643271685,
+      "learning_rate": 0.0008037459283387622,
+      "loss": 0.7265,
+      "step": 1037
+    },
+    {
+      "epoch": 2.281318681318681,
+      "grad_norm": 0.005697134882211685,
+      "learning_rate": 0.0008013029315960912,
+      "loss": 0.8472,
+      "step": 1038
+    },
+    {
+      "epoch": 2.2835164835164834,
+      "grad_norm": 0.013355882838368416,
+      "learning_rate": 0.0007988599348534202,
+      "loss": 0.5887,
+      "step": 1039
+    },
+    {
+      "epoch": 2.2857142857142856,
+      "grad_norm": 0.00827785674482584,
+      "learning_rate": 0.0007964169381107492,
+      "loss": 0.6531,
+      "step": 1040
+    },
+    {
+      "epoch": 2.2879120879120878,
+      "grad_norm": 0.004666945897042751,
+      "learning_rate": 0.0007939739413680782,
+      "loss": 0.9891,
+      "step": 1041
+    },
+    {
+      "epoch": 2.29010989010989,
+      "grad_norm": 0.006005983799695969,
+      "learning_rate": 0.0007915309446254072,
+      "loss": 0.8174,
+      "step": 1042
+    },
+    {
+      "epoch": 2.292307692307692,
+      "grad_norm": 0.005079084075987339,
+      "learning_rate": 0.0007890879478827362,
+      "loss": 0.8898,
+      "step": 1043
+    },
+    {
+      "epoch": 2.2945054945054943,
+      "grad_norm": 0.011428209953010082,
+      "learning_rate": 0.0007866449511400652,
+      "loss": 0.9789,
+      "step": 1044
+    },
+    {
+      "epoch": 2.2967032967032965,
+      "grad_norm": 0.007478225510567427,
+      "learning_rate": 0.0007842019543973941,
+      "loss": 0.809,
+      "step": 1045
+    },
+    {
+      "epoch": 2.2989010989010987,
+      "grad_norm": 0.009910776279866695,
+      "learning_rate": 0.0007817589576547231,
+      "loss": 1.0892,
+      "step": 1046
+    },
+    {
+      "epoch": 2.301098901098901,
+      "grad_norm": 0.010092444717884064,
+      "learning_rate": 0.0007793159609120521,
+      "loss": 0.8665,
+      "step": 1047
+    },
+    {
+      "epoch": 2.303296703296703,
+      "grad_norm": 0.00644384790211916,
+      "learning_rate": 0.0007768729641693811,
+      "loss": 0.8094,
+      "step": 1048
+    },
+    {
+      "epoch": 2.3054945054945053,
+      "grad_norm": 0.007512685377150774,
+      "learning_rate": 0.0007744299674267101,
+      "loss": 0.776,
+      "step": 1049
+    },
+    {
+      "epoch": 2.3076923076923075,
+      "grad_norm": 0.005838407669216394,
+      "learning_rate": 0.0007719869706840391,
+      "loss": 0.7052,
+      "step": 1050
+    },
+    {
+      "epoch": 2.3098901098901097,
+      "grad_norm": 0.010408594273030758,
+      "learning_rate": 0.0007695439739413681,
+      "loss": 0.8556,
+      "step": 1051
+    },
+    {
+      "epoch": 2.312087912087912,
+      "grad_norm": 0.005451490171253681,
+      "learning_rate": 0.0007671009771986971,
+      "loss": 0.6526,
+      "step": 1052
+    },
+    {
+      "epoch": 2.314285714285714,
+      "grad_norm": 0.006650417577475309,
+      "learning_rate": 0.0007646579804560261,
+      "loss": 0.7899,
+      "step": 1053
+    },
+    {
+      "epoch": 2.3164835164835162,
+      "grad_norm": 0.006644013803452253,
+      "learning_rate": 0.0007622149837133551,
+      "loss": 0.8008,
+      "step": 1054
+    },
+    {
+      "epoch": 2.3186813186813184,
+      "grad_norm": 0.00711479689925909,
+      "learning_rate": 0.000759771986970684,
+      "loss": 0.8347,
+      "step": 1055
+    },
+    {
+      "epoch": 2.320879120879121,
+      "grad_norm": 0.008550095371901989,
+      "learning_rate": 0.000757328990228013,
+      "loss": 0.7416,
+      "step": 1056
+    },
+    {
+      "epoch": 2.3230769230769233,
+      "grad_norm": 0.005029153544455767,
+      "learning_rate": 0.000754885993485342,
+      "loss": 0.7632,
+      "step": 1057
+    },
+    {
+      "epoch": 2.3252747252747255,
+      "grad_norm": 0.00951320305466652,
+      "learning_rate": 0.000752442996742671,
+      "loss": 0.8013,
+      "step": 1058
+    },
+    {
+      "epoch": 2.3274725274725276,
+      "grad_norm": 0.00728795537725091,
+      "learning_rate": 0.00075,
+      "loss": 0.694,
+      "step": 1059
+    },
+    {
+      "epoch": 2.32967032967033,
+      "grad_norm": 0.005849994253367186,
+      "learning_rate": 0.000747557003257329,
+      "loss": 0.9843,
+      "step": 1060
+    },
+    {
+      "epoch": 2.331868131868132,
+      "grad_norm": 0.006167943589389324,
+      "learning_rate": 0.000745114006514658,
+      "loss": 0.7919,
+      "step": 1061
+    },
+    {
+      "epoch": 2.334065934065934,
+      "grad_norm": 0.005467480514198542,
+      "learning_rate": 0.000742671009771987,
+      "loss": 0.7186,
+      "step": 1062
+    },
+    {
+      "epoch": 2.3362637362637364,
+      "grad_norm": 0.005549319554120302,
+      "learning_rate": 0.000740228013029316,
+      "loss": 0.9921,
+      "step": 1063
+    },
+    {
+      "epoch": 2.3384615384615386,
+      "grad_norm": 0.004933380056172609,
+      "learning_rate": 0.0007377850162866449,
+      "loss": 0.9189,
+      "step": 1064
+    },
+    {
+      "epoch": 2.340659340659341,
+      "grad_norm": 0.005842259619385004,
+      "learning_rate": 0.0007353420195439739,
+      "loss": 0.679,
+      "step": 1065
+    },
+    {
+      "epoch": 2.342857142857143,
+      "grad_norm": 0.006026857998222113,
+      "learning_rate": 0.0007328990228013029,
+      "loss": 0.8746,
+      "step": 1066
+    },
+    {
+      "epoch": 2.345054945054945,
+      "grad_norm": 0.00629248283803463,
+      "learning_rate": 0.0007304560260586319,
+      "loss": 0.6607,
+      "step": 1067
+    },
+    {
+      "epoch": 2.3472527472527474,
+      "grad_norm": 0.005209072958678007,
+      "learning_rate": 0.0007280130293159609,
+      "loss": 0.7142,
+      "step": 1068
+    },
+    {
+      "epoch": 2.3494505494505495,
+      "grad_norm": 0.009396166540682316,
+      "learning_rate": 0.0007255700325732899,
+      "loss": 0.7975,
+      "step": 1069
+    },
+    {
+      "epoch": 2.3516483516483517,
+      "grad_norm": 0.006473151966929436,
+      "learning_rate": 0.0007231270358306189,
+      "loss": 0.7682,
+      "step": 1070
+    },
+    {
+      "epoch": 2.353846153846154,
+      "grad_norm": 0.004292244557291269,
+      "learning_rate": 0.0007206840390879479,
+      "loss": 0.7448,
+      "step": 1071
+    },
+    {
+      "epoch": 2.356043956043956,
+      "grad_norm": 0.00597755191847682,
+      "learning_rate": 0.0007182410423452769,
+      "loss": 0.7386,
+      "step": 1072
+    },
+    {
+      "epoch": 2.3582417582417583,
+      "grad_norm": 0.0045239003375172615,
+      "learning_rate": 0.0007157980456026059,
+      "loss": 0.9062,
+      "step": 1073
+    },
+    {
+      "epoch": 2.3604395604395605,
+      "grad_norm": 0.008336522616446018,
+      "learning_rate": 0.0007133550488599348,
+      "loss": 0.8071,
+      "step": 1074
+    },
+    {
+      "epoch": 2.3626373626373627,
+      "grad_norm": 0.008937190286815166,
+      "learning_rate": 0.0007109120521172638,
+      "loss": 0.587,
+      "step": 1075
+    },
+    {
+      "epoch": 2.364835164835165,
+      "grad_norm": 0.0033206401858478785,
+      "learning_rate": 0.0007084690553745928,
+      "loss": 0.8011,
+      "step": 1076
+    },
+    {
+      "epoch": 2.367032967032967,
+      "grad_norm": 0.005448639392852783,
+      "learning_rate": 0.0007060260586319218,
+      "loss": 0.6568,
+      "step": 1077
+    },
+    {
+      "epoch": 2.3692307692307693,
+      "grad_norm": 0.004356515593826771,
+      "learning_rate": 0.0007035830618892508,
+      "loss": 0.6866,
+      "step": 1078
+    },
+    {
+      "epoch": 2.3714285714285714,
+      "grad_norm": 0.006029800977557898,
+      "learning_rate": 0.0007011400651465798,
+      "loss": 0.8592,
+      "step": 1079
+    },
+    {
+      "epoch": 2.3736263736263736,
+      "grad_norm": 0.008941984735429287,
+      "learning_rate": 0.0006986970684039088,
+      "loss": 0.7481,
+      "step": 1080
+    },
+    {
+      "epoch": 2.375824175824176,
+      "grad_norm": 0.005498903337866068,
+      "learning_rate": 0.0006962540716612378,
+      "loss": 0.7493,
+      "step": 1081
+    },
+    {
+      "epoch": 2.378021978021978,
+      "grad_norm": 0.007452168967574835,
+      "learning_rate": 0.0006938110749185668,
+      "loss": 0.9569,
+      "step": 1082
+    },
+    {
+      "epoch": 2.38021978021978,
+      "grad_norm": 0.006572681479156017,
+      "learning_rate": 0.0006913680781758957,
+      "loss": 0.8082,
+      "step": 1083
+    },
+    {
+      "epoch": 2.3824175824175824,
+      "grad_norm": 0.007801783736795187,
+      "learning_rate": 0.0006889250814332247,
+      "loss": 0.7759,
+      "step": 1084
+    },
+    {
+      "epoch": 2.3846153846153846,
+      "grad_norm": 0.005979107692837715,
+      "learning_rate": 0.0006864820846905537,
+      "loss": 0.6436,
+      "step": 1085
+    },
+    {
+      "epoch": 2.3868131868131868,
+      "grad_norm": 0.010219307616353035,
+      "learning_rate": 0.0006840390879478827,
+      "loss": 0.9172,
+      "step": 1086
+    },
+    {
+      "epoch": 2.389010989010989,
+      "grad_norm": 0.007844332605600357,
+      "learning_rate": 0.0006815960912052117,
+      "loss": 0.8861,
+      "step": 1087
+    },
+    {
+      "epoch": 2.391208791208791,
+      "grad_norm": 0.007063568104058504,
+      "learning_rate": 0.0006791530944625407,
+      "loss": 1.0757,
+      "step": 1088
+    },
+    {
+      "epoch": 2.3934065934065933,
+      "grad_norm": 0.008408681489527225,
+      "learning_rate": 0.0006767100977198697,
+      "loss": 0.7707,
+      "step": 1089
+    },
+    {
+      "epoch": 2.3956043956043955,
+      "grad_norm": 0.010122616775333881,
+      "learning_rate": 0.0006742671009771987,
+      "loss": 0.7857,
+      "step": 1090
+    },
+    {
+      "epoch": 2.3978021978021977,
+      "grad_norm": 0.011598248966038227,
+      "learning_rate": 0.0006718241042345277,
+      "loss": 1.032,
+      "step": 1091
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.008146912790834904,
+      "learning_rate": 0.0006693811074918567,
+      "loss": 0.7476,
+      "step": 1092
+    },
+    {
+      "epoch": 2.402197802197802,
+      "grad_norm": 0.00666767917573452,
+      "learning_rate": 0.0006669381107491856,
+      "loss": 0.9473,
+      "step": 1093
+    },
+    {
+      "epoch": 2.4043956043956043,
+      "grad_norm": 0.006723317783325911,
+      "learning_rate": 0.0006644951140065146,
+      "loss": 1.0651,
+      "step": 1094
+    },
+    {
+      "epoch": 2.4065934065934065,
+      "grad_norm": 0.006567904260009527,
+      "learning_rate": 0.0006620521172638436,
+      "loss": 0.7228,
+      "step": 1095
+    },
+    {
+      "epoch": 2.4087912087912087,
+      "grad_norm": 0.026447663083672523,
+      "learning_rate": 0.0006596091205211726,
+      "loss": 1.1001,
+      "step": 1096
+    },
+    {
+      "epoch": 2.410989010989011,
+      "grad_norm": 0.005665752105414867,
+      "learning_rate": 0.0006571661237785016,
+      "loss": 0.8601,
+      "step": 1097
+    },
+    {
+      "epoch": 2.413186813186813,
+      "grad_norm": 0.0051101879216730595,
+      "learning_rate": 0.0006547231270358307,
+      "loss": 0.6404,
+      "step": 1098
+    },
+    {
+      "epoch": 2.4153846153846152,
+      "grad_norm": 0.004279910586774349,
+      "learning_rate": 0.0006522801302931597,
+      "loss": 0.7489,
+      "step": 1099
+    },
+    {
+      "epoch": 2.4175824175824174,
+      "grad_norm": 0.0063419826328754425,
+      "learning_rate": 0.0006498371335504886,
+      "loss": 0.7448,
+      "step": 1100
+    },
+    {
+      "epoch": 2.4197802197802196,
+      "grad_norm": 0.005367879755795002,
+      "learning_rate": 0.0006473941368078176,
+      "loss": 0.8237,
+      "step": 1101
+    },
+    {
+      "epoch": 2.421978021978022,
+      "grad_norm": 0.005614032503217459,
+      "learning_rate": 0.0006449511400651466,
+      "loss": 0.8019,
+      "step": 1102
+    },
+    {
+      "epoch": 2.424175824175824,
+      "grad_norm": 0.013202575966715813,
+      "learning_rate": 0.0006425081433224756,
+      "loss": 0.8222,
+      "step": 1103
+    },
+    {
+      "epoch": 2.426373626373626,
+      "grad_norm": 0.00809451099485159,
+      "learning_rate": 0.0006400651465798046,
+      "loss": 0.7727,
+      "step": 1104
+    },
+    {
+      "epoch": 2.4285714285714284,
+      "grad_norm": 0.016084732487797737,
+      "learning_rate": 0.0006376221498371336,
+      "loss": 1.1138,
+      "step": 1105
+    },
+    {
+      "epoch": 2.430769230769231,
+      "grad_norm": 0.005296765826642513,
+      "learning_rate": 0.0006351791530944626,
+      "loss": 0.6478,
+      "step": 1106
+    },
+    {
+      "epoch": 2.432967032967033,
+      "grad_norm": 0.004712815862149,
+      "learning_rate": 0.0006327361563517916,
+      "loss": 0.9725,
+      "step": 1107
+    },
+    {
+      "epoch": 2.4351648351648354,
+      "grad_norm": 0.006297340150922537,
+      "learning_rate": 0.0006302931596091206,
+      "loss": 0.7987,
+      "step": 1108
+    },
+    {
+      "epoch": 2.4373626373626376,
+      "grad_norm": 0.006704499013721943,
+      "learning_rate": 0.0006278501628664496,
+      "loss": 0.6962,
+      "step": 1109
+    },
+    {
+      "epoch": 2.4395604395604398,
+      "grad_norm": 0.0069049205631017685,
+      "learning_rate": 0.0006254071661237785,
+      "loss": 0.8102,
+      "step": 1110
+    },
+    {
+      "epoch": 2.441758241758242,
+      "grad_norm": 0.012242261320352554,
+      "learning_rate": 0.0006229641693811075,
+      "loss": 0.7229,
+      "step": 1111
+    },
+    {
+      "epoch": 2.443956043956044,
+      "grad_norm": 0.006388845853507519,
+      "learning_rate": 0.0006205211726384365,
+      "loss": 0.9646,
+      "step": 1112
+    },
+    {
+      "epoch": 2.4461538461538463,
+      "grad_norm": 0.004925151821225882,
+      "learning_rate": 0.0006180781758957655,
+      "loss": 0.6689,
+      "step": 1113
+    },
+    {
+      "epoch": 2.4483516483516485,
+      "grad_norm": 0.0067703924141824245,
+      "learning_rate": 0.0006156351791530945,
+      "loss": 0.8554,
+      "step": 1114
+    },
+    {
+      "epoch": 2.4505494505494507,
+      "grad_norm": 0.0077737984247505665,
+      "learning_rate": 0.0006131921824104235,
+      "loss": 0.7357,
+      "step": 1115
+    },
+    {
+      "epoch": 2.452747252747253,
+      "grad_norm": 0.004558132961392403,
+      "learning_rate": 0.0006107491856677525,
+      "loss": 0.76,
+      "step": 1116
+    },
+    {
+      "epoch": 2.454945054945055,
+      "grad_norm": 0.0059403530322015285,
+      "learning_rate": 0.0006083061889250815,
+      "loss": 0.7335,
+      "step": 1117
+    },
+    {
+      "epoch": 2.4571428571428573,
+      "grad_norm": 0.0068601886741817,
+      "learning_rate": 0.0006058631921824105,
+      "loss": 0.8267,
+      "step": 1118
+    },
+    {
+      "epoch": 2.4593406593406595,
+      "grad_norm": 0.006239280104637146,
+      "learning_rate": 0.0006034201954397394,
+      "loss": 0.611,
+      "step": 1119
+    },
+    {
+      "epoch": 2.4615384615384617,
+      "grad_norm": 0.010063497349619865,
+      "learning_rate": 0.0006009771986970684,
+      "loss": 0.7799,
+      "step": 1120
+    },
+    {
+      "epoch": 2.463736263736264,
+      "grad_norm": 0.009236813522875309,
+      "learning_rate": 0.0005985342019543974,
+      "loss": 0.6659,
+      "step": 1121
+    },
+    {
+      "epoch": 2.465934065934066,
+      "grad_norm": 0.006913383491337299,
+      "learning_rate": 0.0005960912052117264,
+      "loss": 0.761,
+      "step": 1122
+    },
+    {
+      "epoch": 2.4681318681318682,
+      "grad_norm": 0.007294321898370981,
+      "learning_rate": 0.0005936482084690554,
+      "loss": 0.8346,
+      "step": 1123
+    },
+    {
+      "epoch": 2.4703296703296704,
+      "grad_norm": 0.00751752033829689,
+      "learning_rate": 0.0005912052117263844,
+      "loss": 0.9411,
+      "step": 1124
+    },
+    {
+      "epoch": 2.4725274725274726,
+      "grad_norm": 0.006319766864180565,
+      "learning_rate": 0.0005887622149837134,
+      "loss": 0.7567,
+      "step": 1125
+    },
+    {
+      "epoch": 2.474725274725275,
+      "grad_norm": 0.0069181774742901325,
+      "learning_rate": 0.0005863192182410424,
+      "loss": 0.816,
+      "step": 1126
+    },
+    {
+      "epoch": 2.476923076923077,
+      "grad_norm": 0.006835447158664465,
+      "learning_rate": 0.0005838762214983714,
+      "loss": 0.7565,
+      "step": 1127
+    },
+    {
+      "epoch": 2.479120879120879,
+      "grad_norm": 0.006264280062168837,
+      "learning_rate": 0.0005814332247557004,
+      "loss": 0.842,
+      "step": 1128
+    },
+    {
+      "epoch": 2.4813186813186814,
+      "grad_norm": 0.011438383720815182,
+      "learning_rate": 0.0005789902280130293,
+      "loss": 0.81,
+      "step": 1129
+    },
+    {
+      "epoch": 2.4835164835164836,
+      "grad_norm": 0.004412983078509569,
+      "learning_rate": 0.0005765472312703583,
+      "loss": 0.8676,
+      "step": 1130
+    },
+    {
+      "epoch": 2.4857142857142858,
+      "grad_norm": 0.013808987103402615,
+      "learning_rate": 0.0005741042345276873,
+      "loss": 0.896,
+      "step": 1131
+    },
+    {
+      "epoch": 2.487912087912088,
+      "grad_norm": 0.011984704062342644,
+      "learning_rate": 0.0005716612377850163,
+      "loss": 0.6658,
+      "step": 1132
+    },
+    {
+      "epoch": 2.49010989010989,
+      "grad_norm": 0.006338491570204496,
+      "learning_rate": 0.0005692182410423453,
+      "loss": 0.7138,
+      "step": 1133
+    },
+    {
+      "epoch": 2.4923076923076923,
+      "grad_norm": 0.005150653421878815,
+      "learning_rate": 0.0005667752442996743,
+      "loss": 0.8859,
+      "step": 1134
+    },
+    {
+      "epoch": 2.4945054945054945,
+      "grad_norm": 0.005591288208961487,
+      "learning_rate": 0.0005643322475570033,
+      "loss": 0.6483,
+      "step": 1135
+    },
+    {
+      "epoch": 2.4967032967032967,
+      "grad_norm": 0.0061248186975717545,
+      "learning_rate": 0.0005618892508143323,
+      "loss": 0.723,
+      "step": 1136
+    },
+    {
+      "epoch": 2.498901098901099,
+      "grad_norm": 0.010665234178304672,
+      "learning_rate": 0.0005594462540716613,
+      "loss": 0.6655,
+      "step": 1137
+    },
+    {
+      "epoch": 2.501098901098901,
+      "grad_norm": 0.00933606456965208,
+      "learning_rate": 0.0005570032573289903,
+      "loss": 0.7156,
+      "step": 1138
+    },
+    {
+      "epoch": 2.5032967032967033,
+      "grad_norm": 0.00852033868432045,
+      "learning_rate": 0.0005545602605863192,
+      "loss": 0.8015,
+      "step": 1139
+    },
+    {
+      "epoch": 2.5054945054945055,
+      "grad_norm": 0.011883478611707687,
+      "learning_rate": 0.0005521172638436482,
+      "loss": 0.8437,
+      "step": 1140
+    },
+    {
+      "epoch": 2.5076923076923077,
+      "grad_norm": 0.011336020193994045,
+      "learning_rate": 0.0005496742671009772,
+      "loss": 0.7213,
+      "step": 1141
+    },
+    {
+      "epoch": 2.50989010989011,
+      "grad_norm": 0.005798524711281061,
+      "learning_rate": 0.0005472312703583062,
+      "loss": 1.0301,
+      "step": 1142
+    },
+    {
+      "epoch": 2.512087912087912,
+      "grad_norm": 0.005573587957769632,
+      "learning_rate": 0.0005447882736156352,
+      "loss": 0.7727,
+      "step": 1143
+    },
+    {
+      "epoch": 2.5142857142857142,
+      "grad_norm": 0.007739666849374771,
+      "learning_rate": 0.0005423452768729642,
+      "loss": 0.7776,
+      "step": 1144
+    },
+    {
+      "epoch": 2.5164835164835164,
+      "grad_norm": 0.007505151443183422,
+      "learning_rate": 0.0005399022801302932,
+      "loss": 0.9331,
+      "step": 1145
+    },
+    {
+      "epoch": 2.5186813186813186,
+      "grad_norm": 0.006109148729592562,
+      "learning_rate": 0.0005374592833876222,
+      "loss": 0.5781,
+      "step": 1146
+    },
+    {
+      "epoch": 2.520879120879121,
+      "grad_norm": 0.0045758457854390144,
+      "learning_rate": 0.0005350162866449512,
+      "loss": 0.5759,
+      "step": 1147
+    },
+    {
+      "epoch": 2.523076923076923,
+      "grad_norm": 0.007664409000426531,
+      "learning_rate": 0.0005325732899022801,
+      "loss": 0.7229,
+      "step": 1148
+    },
+    {
+      "epoch": 2.525274725274725,
+      "grad_norm": 0.004711830988526344,
+      "learning_rate": 0.0005301302931596091,
+      "loss": 0.6693,
+      "step": 1149
+    },
+    {
+      "epoch": 2.5274725274725274,
+      "grad_norm": 0.0080286655575037,
+      "learning_rate": 0.0005276872964169381,
+      "loss": 0.9612,
+      "step": 1150
+    },
+    {
+      "epoch": 2.5296703296703296,
+      "grad_norm": 0.00500887306407094,
+      "learning_rate": 0.0005252442996742671,
+      "loss": 0.6958,
+      "step": 1151
+    },
+    {
+      "epoch": 2.5318681318681318,
+      "grad_norm": 0.006061498541384935,
+      "learning_rate": 0.0005228013029315961,
+      "loss": 0.7614,
+      "step": 1152
+    },
+    {
+      "epoch": 2.534065934065934,
+      "grad_norm": 0.011481314897537231,
+      "learning_rate": 0.0005203583061889251,
+      "loss": 0.8245,
+      "step": 1153
+    },
+    {
+      "epoch": 2.536263736263736,
+      "grad_norm": 0.006708142347633839,
+      "learning_rate": 0.0005179153094462541,
+      "loss": 0.8778,
+      "step": 1154
+    },
+    {
+      "epoch": 2.5384615384615383,
+      "grad_norm": 0.010574715211987495,
+      "learning_rate": 0.0005154723127035831,
+      "loss": 0.8859,
+      "step": 1155
+    },
+    {
+      "epoch": 2.5406593406593405,
+      "grad_norm": 0.0457354411482811,
+      "learning_rate": 0.0005130293159609121,
+      "loss": 0.7601,
+      "step": 1156
+    },
+    {
+      "epoch": 2.5428571428571427,
+      "grad_norm": 0.0058629512786865234,
+      "learning_rate": 0.0005105863192182411,
+      "loss": 0.7435,
+      "step": 1157
+    },
+    {
+      "epoch": 2.545054945054945,
+      "grad_norm": 0.007004350423812866,
+      "learning_rate": 0.00050814332247557,
+      "loss": 0.7477,
+      "step": 1158
+    },
+    {
+      "epoch": 2.547252747252747,
+      "grad_norm": 0.00818250048905611,
+      "learning_rate": 0.000505700325732899,
+      "loss": 0.8122,
+      "step": 1159
+    },
+    {
+      "epoch": 2.5494505494505493,
+      "grad_norm": 0.005867977626621723,
+      "learning_rate": 0.000503257328990228,
+      "loss": 0.6487,
+      "step": 1160
+    },
+    {
+      "epoch": 2.5516483516483515,
+      "grad_norm": 0.007405155338346958,
+      "learning_rate": 0.000500814332247557,
+      "loss": 0.6826,
+      "step": 1161
+    },
+    {
+      "epoch": 2.5538461538461537,
+      "grad_norm": 0.012186862528324127,
+      "learning_rate": 0.000498371335504886,
+      "loss": 0.6902,
+      "step": 1162
+    },
+    {
+      "epoch": 2.556043956043956,
+      "grad_norm": 0.007889930158853531,
+      "learning_rate": 0.000495928338762215,
+      "loss": 0.8965,
+      "step": 1163
+    },
+    {
+      "epoch": 2.558241758241758,
+      "grad_norm": 0.008861697278916836,
+      "learning_rate": 0.000493485342019544,
+      "loss": 0.8075,
+      "step": 1164
+    },
+    {
+      "epoch": 2.5604395604395602,
+      "grad_norm": 0.006033893674612045,
+      "learning_rate": 0.000491042345276873,
+      "loss": 0.8522,
+      "step": 1165
+    },
+    {
+      "epoch": 2.5626373626373624,
+      "grad_norm": 0.009476139210164547,
+      "learning_rate": 0.000488599348534202,
+      "loss": 0.8757,
+      "step": 1166
+    },
+    {
+      "epoch": 2.5648351648351646,
+      "grad_norm": 0.005897150374948978,
+      "learning_rate": 0.0004861563517915309,
+      "loss": 0.6994,
+      "step": 1167
+    },
+    {
+      "epoch": 2.567032967032967,
+      "grad_norm": 0.007385133299976587,
+      "learning_rate": 0.0004837133550488599,
+      "loss": 0.7632,
+      "step": 1168
+    },
+    {
+      "epoch": 2.569230769230769,
+      "grad_norm": 0.005341728217899799,
+      "learning_rate": 0.0004812703583061889,
+      "loss": 0.8806,
+      "step": 1169
+    },
+    {
+      "epoch": 2.571428571428571,
+      "grad_norm": 0.010026124306023121,
+      "learning_rate": 0.00047882736156351787,
+      "loss": 0.7551,
+      "step": 1170
+    },
+    {
+      "epoch": 2.5736263736263734,
+      "grad_norm": 0.008241658098995686,
+      "learning_rate": 0.00047638436482084687,
+      "loss": 0.7521,
+      "step": 1171
+    },
+    {
+      "epoch": 2.5758241758241756,
+      "grad_norm": 0.004722557030618191,
+      "learning_rate": 0.00047394136807817587,
+      "loss": 0.7217,
+      "step": 1172
+    },
+    {
+      "epoch": 2.5780219780219777,
+      "grad_norm": 0.005738707259297371,
+      "learning_rate": 0.00047149837133550487,
+      "loss": 0.9659,
+      "step": 1173
+    },
+    {
+      "epoch": 2.5802197802197804,
+      "grad_norm": 0.006637862417846918,
+      "learning_rate": 0.00046905537459283387,
+      "loss": 0.8346,
+      "step": 1174
+    },
+    {
+      "epoch": 2.5824175824175826,
+      "grad_norm": 0.006731619127094746,
+      "learning_rate": 0.0004666123778501629,
+      "loss": 1.0527,
+      "step": 1175
+    },
+    {
+      "epoch": 2.5846153846153848,
+      "grad_norm": 0.010952206328511238,
+      "learning_rate": 0.0004641693811074919,
+      "loss": 0.7793,
+      "step": 1176
+    },
+    {
+      "epoch": 2.586813186813187,
+      "grad_norm": 0.006575143430382013,
+      "learning_rate": 0.00046172638436482087,
+      "loss": 0.9431,
+      "step": 1177
+    },
+    {
+      "epoch": 2.589010989010989,
+      "grad_norm": 0.006040576845407486,
+      "learning_rate": 0.00045928338762214987,
+      "loss": 0.6726,
+      "step": 1178
+    },
+    {
+      "epoch": 2.5912087912087913,
+      "grad_norm": 0.0067054578103125095,
+      "learning_rate": 0.0004568403908794789,
+      "loss": 0.7867,
+      "step": 1179
+    },
+    {
+      "epoch": 2.5934065934065935,
+      "grad_norm": 0.00485964072868228,
+      "learning_rate": 0.0004543973941368079,
+      "loss": 0.7462,
+      "step": 1180
+    },
+    {
+      "epoch": 2.5956043956043957,
+      "grad_norm": 0.007617277558892965,
+      "learning_rate": 0.0004519543973941369,
+      "loss": 0.8472,
+      "step": 1181
+    },
+    {
+      "epoch": 2.597802197802198,
+      "grad_norm": 0.0047485604882240295,
+      "learning_rate": 0.0004495114006514658,
+      "loss": 0.8067,
+      "step": 1182
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.007456649094820023,
+      "learning_rate": 0.0004470684039087948,
+      "loss": 0.7809,
+      "step": 1183
+    },
+    {
+      "epoch": 2.6021978021978023,
+      "grad_norm": 0.005027905106544495,
+      "learning_rate": 0.0004446254071661238,
+      "loss": 0.6813,
+      "step": 1184
+    },
+    {
+      "epoch": 2.6043956043956045,
+      "grad_norm": 0.006729634013026953,
+      "learning_rate": 0.0004421824104234528,
+      "loss": 0.7651,
+      "step": 1185
+    },
+    {
+      "epoch": 2.6065934065934067,
+      "grad_norm": 0.008727509528398514,
+      "learning_rate": 0.0004397394136807818,
+      "loss": 0.8385,
+      "step": 1186
+    },
+    {
+      "epoch": 2.608791208791209,
+      "grad_norm": 0.004165520425885916,
+      "learning_rate": 0.00043729641693811077,
+      "loss": 0.874,
+      "step": 1187
+    },
+    {
+      "epoch": 2.610989010989011,
+      "grad_norm": 0.004646054469048977,
+      "learning_rate": 0.00043485342019543977,
+      "loss": 0.6984,
+      "step": 1188
+    },
+    {
+      "epoch": 2.6131868131868132,
+      "grad_norm": 0.005587347783148289,
+      "learning_rate": 0.00043241042345276877,
+      "loss": 0.8589,
+      "step": 1189
+    },
+    {
+      "epoch": 2.6153846153846154,
+      "grad_norm": 0.0099089490249753,
+      "learning_rate": 0.00042996742671009777,
+      "loss": 0.7562,
+      "step": 1190
+    },
+    {
+      "epoch": 2.6175824175824176,
+      "grad_norm": 0.005145385395735502,
+      "learning_rate": 0.0004275244299674267,
+      "loss": 0.7286,
+      "step": 1191
+    },
+    {
+      "epoch": 2.61978021978022,
+      "grad_norm": 0.011309281922876835,
+      "learning_rate": 0.0004250814332247557,
+      "loss": 0.7878,
+      "step": 1192
+    },
+    {
+      "epoch": 2.621978021978022,
+      "grad_norm": 0.003077469300478697,
+      "learning_rate": 0.0004226384364820847,
+      "loss": 0.7573,
+      "step": 1193
+    },
+    {
+      "epoch": 2.624175824175824,
+      "grad_norm": 0.0060254959389567375,
+      "learning_rate": 0.0004201954397394137,
+      "loss": 0.7069,
+      "step": 1194
+    },
+    {
+      "epoch": 2.6263736263736264,
+      "grad_norm": 0.006287756841629744,
+      "learning_rate": 0.0004177524429967427,
+      "loss": 0.9132,
+      "step": 1195
+    },
+    {
+      "epoch": 2.6285714285714286,
+      "grad_norm": 0.007869000546634197,
+      "learning_rate": 0.00041530944625407166,
+      "loss": 0.6477,
+      "step": 1196
+    },
+    {
+      "epoch": 2.6307692307692307,
+      "grad_norm": 0.005576707888394594,
+      "learning_rate": 0.00041286644951140066,
+      "loss": 0.6314,
+      "step": 1197
+    },
+    {
+      "epoch": 2.632967032967033,
+      "grad_norm": 0.0051623377948999405,
+      "learning_rate": 0.00041042345276872966,
+      "loss": 0.7825,
+      "step": 1198
+    },
+    {
+      "epoch": 2.635164835164835,
+      "grad_norm": 0.007810649927705526,
+      "learning_rate": 0.00040798045602605866,
+      "loss": 0.9354,
+      "step": 1199
+    },
+    {
+      "epoch": 2.6373626373626373,
+      "grad_norm": 0.012851850129663944,
+      "learning_rate": 0.00040553745928338767,
+      "loss": 0.732,
+      "step": 1200
+    },
+    {
+      "epoch": 2.6395604395604395,
+      "grad_norm": 0.00558412866666913,
+      "learning_rate": 0.0004030944625407166,
+      "loss": 0.7945,
+      "step": 1201
+    },
+    {
+      "epoch": 2.6417582417582417,
+      "grad_norm": 0.006981994025409222,
+      "learning_rate": 0.0004006514657980456,
+      "loss": 0.8051,
+      "step": 1202
+    },
+    {
+      "epoch": 2.643956043956044,
+      "grad_norm": 0.00497604813426733,
+      "learning_rate": 0.0003982084690553746,
+      "loss": 0.8238,
+      "step": 1203
+    },
+    {
+      "epoch": 2.646153846153846,
+      "grad_norm": 0.005835412535816431,
+      "learning_rate": 0.0003957654723127036,
+      "loss": 0.6512,
+      "step": 1204
+    },
+    {
+      "epoch": 2.6483516483516483,
+      "grad_norm": 0.007866711355745792,
+      "learning_rate": 0.0003933224755700326,
+      "loss": 0.8699,
+      "step": 1205
+    },
+    {
+      "epoch": 2.6505494505494505,
+      "grad_norm": 0.008231962099671364,
+      "learning_rate": 0.00039087947882736156,
+      "loss": 0.848,
+      "step": 1206
+    },
+    {
+      "epoch": 2.6527472527472526,
+      "grad_norm": 0.006901828572154045,
+      "learning_rate": 0.00038843648208469056,
+      "loss": 0.7485,
+      "step": 1207
+    },
+    {
+      "epoch": 2.654945054945055,
+      "grad_norm": 0.005917402915656567,
+      "learning_rate": 0.00038599348534201956,
+      "loss": 0.7081,
+      "step": 1208
+    },
+    {
+      "epoch": 2.657142857142857,
+      "grad_norm": 0.005372255574911833,
+      "learning_rate": 0.00038355048859934856,
+      "loss": 1.009,
+      "step": 1209
+    },
+    {
+      "epoch": 2.659340659340659,
+      "grad_norm": 0.005883066914975643,
+      "learning_rate": 0.00038110749185667756,
+      "loss": 0.841,
+      "step": 1210
+    },
+    {
+      "epoch": 2.6615384615384614,
+      "grad_norm": 0.007924503646790981,
+      "learning_rate": 0.0003786644951140065,
+      "loss": 0.8647,
+      "step": 1211
+    },
+    {
+      "epoch": 2.6637362637362636,
+      "grad_norm": 0.006905787158757448,
+      "learning_rate": 0.0003762214983713355,
+      "loss": 0.6716,
+      "step": 1212
+    },
+    {
+      "epoch": 2.665934065934066,
+      "grad_norm": 0.005142308305948973,
+      "learning_rate": 0.0003737785016286645,
+      "loss": 0.6832,
+      "step": 1213
+    },
+    {
+      "epoch": 2.668131868131868,
+      "grad_norm": 0.0064476002007722855,
+      "learning_rate": 0.0003713355048859935,
+      "loss": 0.619,
+      "step": 1214
+    },
+    {
+      "epoch": 2.67032967032967,
+      "grad_norm": 0.008479131385684013,
+      "learning_rate": 0.00036889250814332245,
+      "loss": 0.6898,
+      "step": 1215
+    },
+    {
+      "epoch": 2.672527472527473,
+      "grad_norm": 0.006969228852540255,
+      "learning_rate": 0.00036644951140065146,
+      "loss": 0.6664,
+      "step": 1216
+    },
+    {
+      "epoch": 2.674725274725275,
+      "grad_norm": 0.006685610394924879,
+      "learning_rate": 0.00036400651465798046,
+      "loss": 0.6636,
+      "step": 1217
+    },
+    {
+      "epoch": 2.676923076923077,
+      "grad_norm": 0.005391824059188366,
+      "learning_rate": 0.00036156351791530946,
+      "loss": 0.8639,
+      "step": 1218
+    },
+    {
+      "epoch": 2.6791208791208794,
+      "grad_norm": 0.008159205317497253,
+      "learning_rate": 0.00035912052117263846,
+      "loss": 0.6217,
+      "step": 1219
+    },
+    {
+      "epoch": 2.6813186813186816,
+      "grad_norm": 0.005829444620758295,
+      "learning_rate": 0.0003566775244299674,
+      "loss": 0.6635,
+      "step": 1220
+    },
+    {
+      "epoch": 2.6835164835164838,
+      "grad_norm": 0.00569180166348815,
+      "learning_rate": 0.0003542345276872964,
+      "loss": 0.6824,
+      "step": 1221
+    },
+    {
+      "epoch": 2.685714285714286,
+      "grad_norm": 0.006121391896158457,
+      "learning_rate": 0.0003517915309446254,
+      "loss": 0.8428,
+      "step": 1222
+    },
+    {
+      "epoch": 2.687912087912088,
+      "grad_norm": 0.005177916027605534,
+      "learning_rate": 0.0003493485342019544,
+      "loss": 0.8921,
+      "step": 1223
+    },
+    {
+      "epoch": 2.6901098901098903,
+      "grad_norm": 0.004657004959881306,
+      "learning_rate": 0.0003469055374592834,
+      "loss": 0.785,
+      "step": 1224
+    },
+    {
+      "epoch": 2.6923076923076925,
+      "grad_norm": 0.005992937833070755,
+      "learning_rate": 0.00034446254071661235,
+      "loss": 0.8557,
+      "step": 1225
+    },
+    {
+      "epoch": 2.6945054945054947,
+      "grad_norm": 0.006959916558116674,
+      "learning_rate": 0.00034201954397394135,
+      "loss": 0.6442,
+      "step": 1226
+    },
+    {
+      "epoch": 2.696703296703297,
+      "grad_norm": 0.005126854870468378,
+      "learning_rate": 0.00033957654723127035,
+      "loss": 0.6299,
+      "step": 1227
+    },
+    {
+      "epoch": 2.698901098901099,
+      "grad_norm": 0.007516913115978241,
+      "learning_rate": 0.00033713355048859935,
+      "loss": 0.6956,
+      "step": 1228
+    },
+    {
+      "epoch": 2.7010989010989013,
+      "grad_norm": 0.00566484872251749,
+      "learning_rate": 0.00033469055374592835,
+      "loss": 0.9975,
+      "step": 1229
+    },
+    {
+      "epoch": 2.7032967032967035,
+      "grad_norm": 0.005924667231738567,
+      "learning_rate": 0.0003322475570032573,
+      "loss": 0.9001,
+      "step": 1230
+    },
+    {
+      "epoch": 2.7054945054945057,
+      "grad_norm": 0.007858403027057648,
+      "learning_rate": 0.0003298045602605863,
+      "loss": 0.7748,
+      "step": 1231
+    },
+    {
+      "epoch": 2.707692307692308,
+      "grad_norm": 0.00630589947104454,
+      "learning_rate": 0.00032736156351791535,
+      "loss": 0.8235,
+      "step": 1232
+    },
+    {
+      "epoch": 2.70989010989011,
+      "grad_norm": 0.005290319677442312,
+      "learning_rate": 0.0003249185667752443,
+      "loss": 0.8705,
+      "step": 1233
+    },
+    {
+      "epoch": 2.7120879120879122,
+      "grad_norm": 0.009120403788983822,
+      "learning_rate": 0.0003224755700325733,
+      "loss": 0.8405,
+      "step": 1234
+    },
+    {
+      "epoch": 2.7142857142857144,
+      "grad_norm": 0.005391221027821302,
+      "learning_rate": 0.0003200325732899023,
+      "loss": 0.7737,
+      "step": 1235
+    },
+    {
+      "epoch": 2.7164835164835166,
+      "grad_norm": 0.010553315281867981,
+      "learning_rate": 0.0003175895765472313,
+      "loss": 0.6344,
+      "step": 1236
+    },
+    {
+      "epoch": 2.718681318681319,
+      "grad_norm": 0.005718598607927561,
+      "learning_rate": 0.0003151465798045603,
+      "loss": 0.6366,
+      "step": 1237
+    },
+    {
+      "epoch": 2.720879120879121,
+      "grad_norm": 0.0057194409891963005,
+      "learning_rate": 0.00031270358306188925,
+      "loss": 0.8948,
+      "step": 1238
+    },
+    {
+      "epoch": 2.723076923076923,
+      "grad_norm": 0.0050989058800041676,
+      "learning_rate": 0.00031026058631921825,
+      "loss": 0.7703,
+      "step": 1239
+    },
+    {
+      "epoch": 2.7252747252747254,
+      "grad_norm": 0.005610834341496229,
+      "learning_rate": 0.00030781758957654725,
+      "loss": 0.837,
+      "step": 1240
+    },
+    {
+      "epoch": 2.7274725274725276,
+      "grad_norm": 0.00967538170516491,
+      "learning_rate": 0.00030537459283387625,
+      "loss": 0.6747,
+      "step": 1241
+    },
+    {
+      "epoch": 2.7296703296703297,
+      "grad_norm": 0.004189359489828348,
+      "learning_rate": 0.00030293159609120525,
+      "loss": 0.6546,
+      "step": 1242
+    },
+    {
+      "epoch": 2.731868131868132,
+      "grad_norm": 0.008595873601734638,
+      "learning_rate": 0.0003004885993485342,
+      "loss": 0.7935,
+      "step": 1243
+    },
+    {
+      "epoch": 2.734065934065934,
+      "grad_norm": 0.006577143445611,
+      "learning_rate": 0.0002980456026058632,
+      "loss": 0.8892,
+      "step": 1244
+    },
+    {
+      "epoch": 2.7362637362637363,
+      "grad_norm": 0.0073424819856882095,
+      "learning_rate": 0.0002956026058631922,
+      "loss": 0.8213,
+      "step": 1245
+    },
+    {
+      "epoch": 2.7384615384615385,
+      "grad_norm": 0.007525932509452105,
+      "learning_rate": 0.0002931596091205212,
+      "loss": 0.7586,
+      "step": 1246
+    },
+    {
+      "epoch": 2.7406593406593407,
+      "grad_norm": 0.005765283014625311,
+      "learning_rate": 0.0002907166123778502,
+      "loss": 0.7407,
+      "step": 1247
+    },
+    {
+      "epoch": 2.742857142857143,
+      "grad_norm": 0.0072677768766880035,
+      "learning_rate": 0.00028827361563517914,
+      "loss": 0.773,
+      "step": 1248
+    },
+    {
+      "epoch": 2.745054945054945,
+      "grad_norm": 0.008345763199031353,
+      "learning_rate": 0.00028583061889250814,
+      "loss": 0.858,
+      "step": 1249
+    },
+    {
+      "epoch": 2.7472527472527473,
+      "grad_norm": 0.011663106270134449,
+      "learning_rate": 0.00028338762214983714,
+      "loss": 0.6223,
+      "step": 1250
+    },
+    {
+      "epoch": 2.7494505494505495,
+      "grad_norm": 0.03351084515452385,
+      "learning_rate": 0.00028094462540716614,
+      "loss": 0.7555,
+      "step": 1251
+    },
+    {
+      "epoch": 2.7516483516483516,
+      "grad_norm": 0.004764399025589228,
+      "learning_rate": 0.00027850162866449514,
+      "loss": 0.8833,
+      "step": 1252
+    },
+    {
+      "epoch": 2.753846153846154,
+      "grad_norm": 0.004721418954432011,
+      "learning_rate": 0.0002760586319218241,
+      "loss": 0.915,
+      "step": 1253
+    },
+    {
+      "epoch": 2.756043956043956,
+      "grad_norm": 0.0055750347673892975,
+      "learning_rate": 0.0002736156351791531,
+      "loss": 0.661,
+      "step": 1254
+    },
+    {
+      "epoch": 2.758241758241758,
+      "grad_norm": 0.020436737686395645,
+      "learning_rate": 0.0002711726384364821,
+      "loss": 0.9074,
+      "step": 1255
+    },
+    {
+      "epoch": 2.7604395604395604,
+      "grad_norm": 0.0050248755142092705,
+      "learning_rate": 0.0002687296416938111,
+      "loss": 0.6571,
+      "step": 1256
+    },
+    {
+      "epoch": 2.7626373626373626,
+      "grad_norm": 0.005791441071778536,
+      "learning_rate": 0.00026628664495114004,
+      "loss": 0.7904,
+      "step": 1257
+    },
+    {
+      "epoch": 2.764835164835165,
+      "grad_norm": 0.005430641118437052,
+      "learning_rate": 0.00026384364820846904,
+      "loss": 0.6995,
+      "step": 1258
+    },
+    {
+      "epoch": 2.767032967032967,
+      "grad_norm": 0.008256346918642521,
+      "learning_rate": 0.00026140065146579804,
+      "loss": 0.6587,
+      "step": 1259
+    },
+    {
+      "epoch": 2.769230769230769,
+      "grad_norm": 0.0068228477612137794,
+      "learning_rate": 0.00025895765472312704,
+      "loss": 0.8348,
+      "step": 1260
+    },
+    {
+      "epoch": 2.7714285714285714,
+      "grad_norm": 0.004999696277081966,
+      "learning_rate": 0.00025651465798045604,
+      "loss": 0.9725,
+      "step": 1261
+    },
+    {
+      "epoch": 2.7736263736263735,
+      "grad_norm": 0.008323190733790398,
+      "learning_rate": 0.000254071661237785,
+      "loss": 0.6829,
+      "step": 1262
+    },
+    {
+      "epoch": 2.7758241758241757,
+      "grad_norm": 0.00417680200189352,
+      "learning_rate": 0.000251628664495114,
+      "loss": 0.8848,
+      "step": 1263
+    },
+    {
+      "epoch": 2.778021978021978,
+      "grad_norm": 0.0048073576763272285,
+      "learning_rate": 0.000249185667752443,
+      "loss": 0.7074,
+      "step": 1264
+    },
+    {
+      "epoch": 2.78021978021978,
+      "grad_norm": 0.008036723360419273,
+      "learning_rate": 0.000246742671009772,
+      "loss": 0.7762,
+      "step": 1265
+    },
+    {
+      "epoch": 2.7824175824175823,
+      "grad_norm": 0.006922819651663303,
+      "learning_rate": 0.000244299674267101,
+      "loss": 0.7655,
+      "step": 1266
+    },
+    {
+      "epoch": 2.7846153846153845,
+      "grad_norm": 0.00796543899923563,
+      "learning_rate": 0.00024185667752442996,
+      "loss": 0.7266,
+      "step": 1267
+    },
+    {
+      "epoch": 2.7868131868131867,
+      "grad_norm": 0.003956203814595938,
+      "learning_rate": 0.00023941368078175893,
+      "loss": 0.8226,
+      "step": 1268
+    },
+    {
+      "epoch": 2.789010989010989,
+      "grad_norm": 0.00589255103841424,
+      "learning_rate": 0.00023697068403908794,
+      "loss": 0.8065,
+      "step": 1269
+    },
+    {
+      "epoch": 2.791208791208791,
+      "grad_norm": 0.0044288006611168385,
+      "learning_rate": 0.00023452768729641694,
+      "loss": 0.792,
+      "step": 1270
+    },
+    {
+      "epoch": 2.7934065934065933,
+      "grad_norm": 0.00599051034078002,
+      "learning_rate": 0.00023208469055374596,
+      "loss": 0.9386,
+      "step": 1271
+    },
+    {
+      "epoch": 2.7956043956043954,
+      "grad_norm": 0.011603651568293571,
+      "learning_rate": 0.00022964169381107494,
+      "loss": 0.8189,
+      "step": 1272
+    },
+    {
+      "epoch": 2.7978021978021976,
+      "grad_norm": 0.007274969015270472,
+      "learning_rate": 0.00022719869706840394,
+      "loss": 0.758,
+      "step": 1273
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.005429944489151239,
+      "learning_rate": 0.0002247557003257329,
+      "loss": 0.6923,
+      "step": 1274
+    },
+    {
+      "epoch": 2.802197802197802,
+      "grad_norm": 0.006422122474759817,
+      "learning_rate": 0.0002223127035830619,
+      "loss": 0.853,
+      "step": 1275
+    },
+    {
+      "epoch": 2.804395604395604,
+      "grad_norm": 0.010066731832921505,
+      "learning_rate": 0.0002198697068403909,
+      "loss": 0.6671,
+      "step": 1276
+    },
+    {
+      "epoch": 2.8065934065934064,
+      "grad_norm": 0.009736527688801289,
+      "learning_rate": 0.00021742671009771988,
+      "loss": 0.9087,
+      "step": 1277
+    },
+    {
+      "epoch": 2.8087912087912086,
+      "grad_norm": 0.004498061258345842,
+      "learning_rate": 0.00021498371335504888,
+      "loss": 0.7966,
+      "step": 1278
+    },
+    {
+      "epoch": 2.8109890109890108,
+      "grad_norm": 0.006300424225628376,
+      "learning_rate": 0.00021254071661237786,
+      "loss": 0.7721,
+      "step": 1279
+    },
+    {
+      "epoch": 2.813186813186813,
+      "grad_norm": 0.008078373037278652,
+      "learning_rate": 0.00021009771986970686,
+      "loss": 0.699,
+      "step": 1280
+    },
+    {
+      "epoch": 2.815384615384615,
+      "grad_norm": 0.00559539720416069,
+      "learning_rate": 0.00020765472312703583,
+      "loss": 0.5677,
+      "step": 1281
+    },
+    {
+      "epoch": 2.8175824175824173,
+      "grad_norm": 0.008244226686656475,
+      "learning_rate": 0.00020521172638436483,
+      "loss": 0.9385,
+      "step": 1282
+    },
+    {
+      "epoch": 2.8197802197802195,
+      "grad_norm": 0.005011278670281172,
+      "learning_rate": 0.00020276872964169383,
+      "loss": 0.7487,
+      "step": 1283
+    },
+    {
+      "epoch": 2.8219780219780217,
+      "grad_norm": 0.018438560888171196,
+      "learning_rate": 0.0002003257328990228,
+      "loss": 0.8418,
+      "step": 1284
+    },
+    {
+      "epoch": 2.824175824175824,
+      "grad_norm": 0.006285513751208782,
+      "learning_rate": 0.0001978827361563518,
+      "loss": 0.677,
+      "step": 1285
+    },
+    {
+      "epoch": 2.826373626373626,
+      "grad_norm": 0.00687903817743063,
+      "learning_rate": 0.00019543973941368078,
+      "loss": 0.7501,
+      "step": 1286
+    },
+    {
+      "epoch": 2.8285714285714287,
+      "grad_norm": 0.005045247729867697,
+      "learning_rate": 0.00019299674267100978,
+      "loss": 0.5691,
+      "step": 1287
+    },
+    {
+      "epoch": 2.830769230769231,
+      "grad_norm": 0.010209862142801285,
+      "learning_rate": 0.00019055374592833878,
+      "loss": 0.7588,
+      "step": 1288
+    },
+    {
+      "epoch": 2.832967032967033,
+      "grad_norm": 0.005651464685797691,
+      "learning_rate": 0.00018811074918566775,
+      "loss": 0.8508,
+      "step": 1289
+    },
+    {
+      "epoch": 2.8351648351648353,
+      "grad_norm": 0.00873302761465311,
+      "learning_rate": 0.00018566775244299675,
+      "loss": 0.8333,
+      "step": 1290
+    },
+    {
+      "epoch": 2.8373626373626375,
+      "grad_norm": 0.010436766780912876,
+      "learning_rate": 0.00018322475570032573,
+      "loss": 0.8147,
+      "step": 1291
+    },
+    {
+      "epoch": 2.8395604395604397,
+      "grad_norm": 0.009002077393233776,
+      "learning_rate": 0.00018078175895765473,
+      "loss": 0.6796,
+      "step": 1292
+    },
+    {
+      "epoch": 2.841758241758242,
+      "grad_norm": 0.009873115457594395,
+      "learning_rate": 0.0001783387622149837,
+      "loss": 0.8289,
+      "step": 1293
+    },
+    {
+      "epoch": 2.843956043956044,
+      "grad_norm": 0.005910174455493689,
+      "learning_rate": 0.0001758957654723127,
+      "loss": 0.8065,
+      "step": 1294
+    },
+    {
+      "epoch": 2.8461538461538463,
+      "grad_norm": 0.004359032958745956,
+      "learning_rate": 0.0001734527687296417,
+      "loss": 0.7094,
+      "step": 1295
+    },
+    {
+      "epoch": 2.8483516483516484,
+      "grad_norm": 0.005306289531290531,
+      "learning_rate": 0.00017100977198697068,
+      "loss": 0.6685,
+      "step": 1296
+    },
+    {
+      "epoch": 2.8505494505494506,
+      "grad_norm": 0.005530740600079298,
+      "learning_rate": 0.00016856677524429968,
+      "loss": 0.6538,
+      "step": 1297
+    },
+    {
+      "epoch": 2.852747252747253,
+      "grad_norm": 0.007298593875020742,
+      "learning_rate": 0.00016612377850162865,
+      "loss": 0.8683,
+      "step": 1298
+    },
+    {
+      "epoch": 2.854945054945055,
+      "grad_norm": 0.005898923613131046,
+      "learning_rate": 0.00016368078175895768,
+      "loss": 1.0714,
+      "step": 1299
+    },
+    {
+      "epoch": 2.857142857142857,
+      "grad_norm": 0.005482466425746679,
+      "learning_rate": 0.00016123778501628665,
+      "loss": 0.6551,
+      "step": 1300
+    },
+    {
+      "epoch": 2.8593406593406594,
+      "grad_norm": 0.0067042275331914425,
+      "learning_rate": 0.00015879478827361565,
+      "loss": 0.9687,
+      "step": 1301
+    },
+    {
+      "epoch": 2.8615384615384616,
+      "grad_norm": 0.010804589837789536,
+      "learning_rate": 0.00015635179153094462,
+      "loss": 0.8882,
+      "step": 1302
+    },
+    {
+      "epoch": 2.8637362637362638,
+      "grad_norm": 0.008368220180273056,
+      "learning_rate": 0.00015390879478827362,
+      "loss": 0.71,
+      "step": 1303
+    },
+    {
+      "epoch": 2.865934065934066,
+      "grad_norm": 0.006030814256519079,
+      "learning_rate": 0.00015146579804560262,
+      "loss": 0.6942,
+      "step": 1304
+    },
+    {
+      "epoch": 2.868131868131868,
+      "grad_norm": 0.005499513354152441,
+      "learning_rate": 0.0001490228013029316,
+      "loss": 0.6968,
+      "step": 1305
+    },
+    {
+      "epoch": 2.8703296703296703,
+      "grad_norm": 0.00789455883204937,
+      "learning_rate": 0.0001465798045602606,
+      "loss": 0.9392,
+      "step": 1306
+    },
+    {
+      "epoch": 2.8725274725274725,
+      "grad_norm": 0.017804140225052834,
+      "learning_rate": 0.00014413680781758957,
+      "loss": 0.7391,
+      "step": 1307
+    },
+    {
+      "epoch": 2.8747252747252747,
+      "grad_norm": 0.00666830176487565,
+      "learning_rate": 0.00014169381107491857,
+      "loss": 1.0632,
+      "step": 1308
+    },
+    {
+      "epoch": 2.876923076923077,
+      "grad_norm": 0.010356319136917591,
+      "learning_rate": 0.00013925081433224757,
+      "loss": 0.9907,
+      "step": 1309
+    },
+    {
+      "epoch": 2.879120879120879,
+      "grad_norm": 0.009787401184439659,
+      "learning_rate": 0.00013680781758957655,
+      "loss": 0.6988,
+      "step": 1310
+    },
+    {
+      "epoch": 2.8813186813186813,
+      "grad_norm": 0.0053601390682160854,
+      "learning_rate": 0.00013436482084690555,
+      "loss": 0.8133,
+      "step": 1311
+    },
+    {
+      "epoch": 2.8835164835164835,
+      "grad_norm": 0.007845701649785042,
+      "learning_rate": 0.00013192182410423452,
+      "loss": 0.8405,
+      "step": 1312
+    },
+    {
+      "epoch": 2.8857142857142857,
+      "grad_norm": 0.004617604892700911,
+      "learning_rate": 0.00012947882736156352,
+      "loss": 0.7108,
+      "step": 1313
+    },
+    {
+      "epoch": 2.887912087912088,
+      "grad_norm": 0.004257794935256243,
+      "learning_rate": 0.0001270358306188925,
+      "loss": 0.6768,
+      "step": 1314
+    },
+    {
+      "epoch": 2.89010989010989,
+      "grad_norm": 0.006338286679238081,
+      "learning_rate": 0.0001245928338762215,
+      "loss": 0.805,
+      "step": 1315
+    },
+    {
+      "epoch": 2.8923076923076922,
+      "grad_norm": 0.004883977118879557,
+      "learning_rate": 0.0001221498371335505,
+      "loss": 0.7925,
+      "step": 1316
+    },
+    {
+      "epoch": 2.8945054945054944,
+      "grad_norm": 0.01185915432870388,
+      "learning_rate": 0.00011970684039087947,
+      "loss": 0.638,
+      "step": 1317
+    },
+    {
+      "epoch": 2.8967032967032966,
+      "grad_norm": 0.005129642318934202,
+      "learning_rate": 0.00011726384364820847,
+      "loss": 0.7536,
+      "step": 1318
+    },
+    {
+      "epoch": 2.898901098901099,
+      "grad_norm": 0.0053817518055438995,
+      "learning_rate": 0.00011482084690553747,
+      "loss": 0.628,
+      "step": 1319
+    },
+    {
+      "epoch": 2.901098901098901,
+      "grad_norm": 0.008170418441295624,
+      "learning_rate": 0.00011237785016286646,
+      "loss": 0.6872,
+      "step": 1320
+    },
+    {
+      "epoch": 2.903296703296703,
+      "grad_norm": 0.010877996683120728,
+      "learning_rate": 0.00010993485342019546,
+      "loss": 0.7638,
+      "step": 1321
+    },
+    {
+      "epoch": 2.9054945054945054,
+      "grad_norm": 0.010428745299577713,
+      "learning_rate": 0.00010749185667752444,
+      "loss": 0.8925,
+      "step": 1322
+    },
+    {
+      "epoch": 2.9076923076923076,
+      "grad_norm": 0.015973428264260292,
+      "learning_rate": 0.00010504885993485343,
+      "loss": 0.7875,
+      "step": 1323
+    },
+    {
+      "epoch": 2.9098901098901098,
+      "grad_norm": 0.012647329829633236,
+      "learning_rate": 0.00010260586319218242,
+      "loss": 0.8337,
+      "step": 1324
+    },
+    {
+      "epoch": 2.912087912087912,
+      "grad_norm": 0.006094892974942923,
+      "learning_rate": 0.0001001628664495114,
+      "loss": 0.8218,
+      "step": 1325
+    },
+    {
+      "epoch": 2.914285714285714,
+      "grad_norm": 0.004787009675055742,
+      "learning_rate": 9.771986970684039e-05,
+      "loss": 0.7153,
+      "step": 1326
+    },
+    {
+      "epoch": 2.9164835164835163,
+      "grad_norm": 0.007086843717843294,
+      "learning_rate": 9.527687296416939e-05,
+      "loss": 0.665,
+      "step": 1327
+    },
+    {
+      "epoch": 2.9186813186813185,
+      "grad_norm": 0.006821499206125736,
+      "learning_rate": 9.283387622149838e-05,
+      "loss": 0.6837,
+      "step": 1328
+    },
+    {
+      "epoch": 2.9208791208791207,
+      "grad_norm": 0.007143167778849602,
+      "learning_rate": 9.039087947882736e-05,
+      "loss": 0.8092,
+      "step": 1329
+    },
+    {
+      "epoch": 2.9230769230769234,
+      "grad_norm": 0.007236589677631855,
+      "learning_rate": 8.794788273615635e-05,
+      "loss": 1.0073,
+      "step": 1330
+    },
+    {
+      "epoch": 2.9252747252747255,
+      "grad_norm": 0.004453645553439856,
+      "learning_rate": 8.550488599348534e-05,
+      "loss": 0.7209,
+      "step": 1331
+    },
+    {
+      "epoch": 2.9274725274725277,
+      "grad_norm": 0.006358393933624029,
+      "learning_rate": 8.306188925081432e-05,
+      "loss": 0.7742,
+      "step": 1332
+    },
+    {
+      "epoch": 2.92967032967033,
+      "grad_norm": 0.011724279262125492,
+      "learning_rate": 8.061889250814332e-05,
+      "loss": 0.7391,
+      "step": 1333
+    },
+    {
+      "epoch": 2.931868131868132,
+      "grad_norm": 0.006041393149644136,
+      "learning_rate": 7.817589576547231e-05,
+      "loss": 0.7786,
+      "step": 1334
+    },
+    {
+      "epoch": 2.9340659340659343,
+      "grad_norm": 0.0065650781616568565,
+      "learning_rate": 7.573289902280131e-05,
+      "loss": 0.7948,
+      "step": 1335
+    },
+    {
+      "epoch": 2.9362637362637365,
+      "grad_norm": 0.00551111763343215,
+      "learning_rate": 7.32899022801303e-05,
+      "loss": 0.5948,
+      "step": 1336
+    },
+    {
+      "epoch": 2.9384615384615387,
+      "grad_norm": 0.006154465954750776,
+      "learning_rate": 7.084690553745929e-05,
+      "loss": 0.7422,
+      "step": 1337
+    },
+    {
+      "epoch": 2.940659340659341,
+      "grad_norm": 0.00700214272364974,
+      "learning_rate": 6.840390879478827e-05,
+      "loss": 0.8672,
+      "step": 1338
+    },
+    {
+      "epoch": 2.942857142857143,
+      "grad_norm": 0.004223939031362534,
+      "learning_rate": 6.596091205211726e-05,
+      "loss": 0.8398,
+      "step": 1339
+    },
+    {
+      "epoch": 2.9450549450549453,
+      "grad_norm": 0.005142378155142069,
+      "learning_rate": 6.351791530944625e-05,
+      "loss": 0.783,
+      "step": 1340
+    },
+    {
+      "epoch": 2.9472527472527474,
+      "grad_norm": 0.009445125237107277,
+      "learning_rate": 6.107491856677525e-05,
+      "loss": 0.8301,
+      "step": 1341
+    },
+    {
+      "epoch": 2.9494505494505496,
+      "grad_norm": 0.011691125109791756,
+      "learning_rate": 5.8631921824104234e-05,
+      "loss": 0.8472,
+      "step": 1342
+    },
+    {
+      "epoch": 2.951648351648352,
+      "grad_norm": 0.006737316958606243,
+      "learning_rate": 5.618892508143323e-05,
+      "loss": 0.93,
+      "step": 1343
+    },
+    {
+      "epoch": 2.953846153846154,
+      "grad_norm": 0.006278462707996368,
+      "learning_rate": 5.374592833876222e-05,
+      "loss": 0.8106,
+      "step": 1344
+    },
+    {
+      "epoch": 2.956043956043956,
+      "grad_norm": 0.007377979811280966,
+      "learning_rate": 5.130293159609121e-05,
+      "loss": 0.6747,
+      "step": 1345
+    },
+    {
+      "epoch": 2.9582417582417584,
+      "grad_norm": 0.007514790631830692,
+      "learning_rate": 4.8859934853420195e-05,
+      "loss": 0.9714,
+      "step": 1346
+    },
+    {
+      "epoch": 2.9604395604395606,
+      "grad_norm": 0.009186286479234695,
+      "learning_rate": 4.641693811074919e-05,
+      "loss": 0.868,
+      "step": 1347
+    },
+    {
+      "epoch": 2.9626373626373628,
+      "grad_norm": 0.006162294186651707,
+      "learning_rate": 4.3973941368078175e-05,
+      "loss": 0.8553,
+      "step": 1348
+    },
+    {
+      "epoch": 2.964835164835165,
+      "grad_norm": 0.0074793491512537,
+      "learning_rate": 4.153094462540716e-05,
+      "loss": 0.6576,
+      "step": 1349
+    },
+    {
+      "epoch": 2.967032967032967,
+      "grad_norm": 0.004487420432269573,
+      "learning_rate": 3.9087947882736156e-05,
+      "loss": 0.792,
+      "step": 1350
+    },
+    {
+      "epoch": 2.9692307692307693,
+      "grad_norm": 0.018422601744532585,
+      "learning_rate": 3.664495114006515e-05,
+      "loss": 0.9509,
+      "step": 1351
+    },
+    {
+      "epoch": 2.9714285714285715,
+      "grad_norm": 0.007363209500908852,
+      "learning_rate": 3.4201954397394136e-05,
+      "loss": 0.6394,
+      "step": 1352
+    },
+    {
+      "epoch": 2.9736263736263737,
+      "grad_norm": 0.008500104770064354,
+      "learning_rate": 3.175895765472312e-05,
+      "loss": 0.8714,
+      "step": 1353
+    },
+    {
+      "epoch": 2.975824175824176,
+      "grad_norm": 0.00453991862013936,
+      "learning_rate": 2.9315960912052117e-05,
+      "loss": 0.7767,
+      "step": 1354
+    },
+    {
+      "epoch": 2.978021978021978,
+      "grad_norm": 0.007319234311580658,
+      "learning_rate": 2.687296416938111e-05,
+      "loss": 0.8375,
+      "step": 1355
+    },
+    {
+      "epoch": 2.9802197802197803,
+      "grad_norm": 0.006569344084709883,
+      "learning_rate": 2.4429967426710097e-05,
+      "loss": 0.7974,
+      "step": 1356
+    },
+    {
+      "epoch": 2.9824175824175825,
+      "grad_norm": 0.009709206409752369,
+      "learning_rate": 2.1986970684039088e-05,
+      "loss": 0.8509,
+      "step": 1357
+    },
+    {
+      "epoch": 2.9846153846153847,
+      "grad_norm": 0.015746966004371643,
+      "learning_rate": 1.9543973941368078e-05,
+      "loss": 0.7285,
+      "step": 1358
+    },
+    {
+      "epoch": 2.986813186813187,
+      "grad_norm": 0.005030226893723011,
+      "learning_rate": 1.7100977198697068e-05,
+      "loss": 0.6889,
+      "step": 1359
+    },
+    {
+      "epoch": 2.989010989010989,
+      "grad_norm": 0.004636506550014019,
+      "learning_rate": 1.4657980456026058e-05,
+      "loss": 0.772,
+      "step": 1360
+    },
+    {
+      "epoch": 2.9912087912087912,
+      "grad_norm": 0.00714351050555706,
+      "learning_rate": 1.2214983713355049e-05,
+      "loss": 0.6707,
+      "step": 1361
+    },
+    {
+      "epoch": 2.9934065934065934,
+      "grad_norm": 0.004920687526464462,
+      "learning_rate": 9.771986970684039e-06,
+      "loss": 0.8006,
+      "step": 1362
+    },
+    {
+      "epoch": 2.9956043956043956,
+      "grad_norm": 0.0055404044687747955,
+      "learning_rate": 7.328990228013029e-06,
+      "loss": 0.82,
+      "step": 1363
+    },
+    {
+      "epoch": 2.997802197802198,
+      "grad_norm": 0.004729744046926498,
+      "learning_rate": 4.8859934853420195e-06,
+      "loss": 0.7325,
+      "step": 1364
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.00699635362252593,
+      "learning_rate": 2.4429967426710097e-06,
+      "loss": 0.7439,
+      "step": 1365
+    },
+    {
+      "epoch": 3.0,
+      "step": 1365,
+      "total_flos": 1.0349985918458266e+18,
+      "train_loss": 0.8309022970688649,
+      "train_runtime": 4996.2599,
+      "train_samples_per_second": 4.371,
+      "train_steps_per_second": 0.273
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 1365,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.0349985918458266e+18,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}